Academic Editor: Kazushi Watanabe
Objectives: Foetal surgery is a relatively new branch of medicine; the health providers involved are currently wondering what kind of anaesthesia should be provided to the foetuses. In the last few years, new advances have been reported on foetal sensoriality and capability for feeling pain; meanwhile alerts have been issued on the risks of prolonged anaesthesia in the early infancy. Aim of this paper is reviewing the main data on foetal pain, to be aware of which is the time in pregnancy when it is likely to be felt. The secondary aim is to point out which are the anaesthetics and analgesics appropriate for prenatal surgery. Mechanism: A review of the literature published in the last 20 years in the field of fetal sensoriality and fetal sergery has been carried out; the most pertinent papers have been retrieved, and their conclusions are here summarized and analysed. Findings in brief: Pain can be felt by the human fetus in the second half of pregnancy: data of physiological and behavioural studies show it with increasing evidence, as long as the gestational age increases. With regard to the best anaesthesia in this case, it seems that the mere anaesthetics given to the mother, though apparently sufficient during laparoscopic surgery, are not sufficient to anesthetize the foetus during open surgery; here some references are given for the best and safest foetal direct analgesic treatment. Conclusions: Surgeons should approach carefully fetal surgery, according with the latest findings in this field.
Foetal pain is a problem usually dealt with for its ethical implications; but it is also a problem for the surgeons, who are increasingly asked to perform challenging interventions on the foetus before birth. As this type of surgery has become today quite widespread, it is mandatory having certainty about the type of anaesthetic to be administered to the mother or foetus: this now has become a necessary topic in the modern approach to pain [1, 2]. Aim of this review is to report the state of the art about foetal pain, the opportunity about using anaesthetics in these interventions, and the type of analgesia to be used during prenatal foetal interventions.
For a potentially painful stimulus to generate pain, four conditions are needed. (A) Presence of pain receptors; (B) mediators of the painful stimulus; (C) connection between the receptor and the brain centres that “feel” the pain; (D) effective pain centres (Fig. 1).
Pain transmission and effects in midgestation. The cascade of pain starting at midgestation. From the cause to the signs.
Pain receptors appear in the dermis at about 8 weeks of gestational age (WGA) and in the mucosae at 10 WGA; they are mature at 20–23 weeks, when they become functional for their connection with the brain [3]. These receptors have nonetheless the same density [4, 5] or even a greater density [6] than in adults.
Substance P has been found in the spine cord as early as 8–10 WGA [7, 8] and encephalin at 12–14 weeks GA [7, 9]. The cells that produce endorphins have been detected in the pituitary gland at 12 WGA and their functional maturity is evident at 20 WGA, when they respond to corticotropin-releasing factor stimulation [10].
The development of pain pathways begins early in the foetus. It is composed of four stages. The first appears in the embryo (6th WGA), where the dorsal horn cells of the spinal cord form synapses with the primordial sensory neurons, that (stage 2) attain the cutis of the limbs, of the rest of the body and of the mucosae at, respectively, 11 [11, 12], 15 and 20 WGA [13, 14]. Stage 3 is the migration and differentiation from the spinal cord of the neurons that will form the spino-thalamic tract; this takes place in the rat at 12 days of embryonic life [15, 16], comparable with a human embryo of 30 days [17].
The thalamus begins its maturation around 12 weeks [18] and is complete at around 21 weeks [19, 20]. The amygdala appears at 12 weeks of gestation and is developed at term [21, 22].
Thalamo-cortical fibres appear at 15–23 WGA [22, 23, 24] and attain the subplate at 24 WGA, to progress later to the definitive cortex at 24–32 WGA [25]. This enables the cortex to respond to peripherical stimulations [26]. For other authors, the first subplate projections from the thalamus arrive between weeks 12 and 18 [27, 28], the thalamic afferents begin to reach the somatosensory subplate at 20 WGA. The connections from the thalamus arrive to the subplate as early as 12–18 WGA [27, 28], then they head the somatosensory and the visual subplate at 20 and 20–22 WGA respectively [29, 30], to progress later to the definitive cortex [31].
Mode Network are present since the 35th WGA [32, 33]; this Network has an important role in the development of consciousness. Nonetheless, even before that, another important brain structure is involved with pain processing: the subplate. It appears early (11th WGA) in human foetuses, becomes prominent at 22–34 WGA, then gradually disappears almost totally at 6 months of postnatal life [27], though some subplate cells are still present in adults [34]. The subplate is a transitory primordial cortex where neurons remain a while until their definitive migration to the mature cortex. It is a carrefour for thalamo-cortical fibers, where the information from the thalamus can be processed [35, 36, 37]; it also has a spontaneous electrical activity [38] and has an important role in early human behaviours [39, 40].
Neurons from the subplate are driven to the cortex simultaneously with the arrive of direct fibres from the thalamus to the cortex, starting at around 24 WGA [41]. Then this process increases in speed until the subplate definitively disappears and the cortex is fully developed.
According to some researchers, pain can only be perceived if the patient has a mature and safe brain cortex; nonetheless, others do not agree [42, 43, 44, 45, 46, 47, 48, 49, 50, 51], arguing that the mere presence of the periphery-thalamus connection is enough to let pain be perceived. These latter researchers base their argument on the cases of babies with anencephaly. These babies in fact show an unexpected discriminative awareness [50, 51]: they disclose an apparent capacity of discriminating people, situations, environments, and are capable of visual orientation and musical preferences. As a matter of fact, several stimuli can be processed without the help of the brain cortex [43, 44, 49] and provide useful visual information, or trigger complex experiences such as fear [45, 52]. This can be true also for pain processing, according to some authors [53, 54]: anencephalic foetuses have a withdrawal reaction to pain [55] and similar reactions are described in babies with extensive brain damage [56].
Babies born at 26–31 weeks exhibit coordinated facial expressions in response to heel prick, although these are immature compared to older babies [57]. A behaviour similar to infantile crying has been reported in foetuses using 4D-US scan technique [58] and blink-startle reflex in response to a sudden stimulus has been described in 30-week old foetuses [59]. Some wonder whether these reactions may constitute pain or just a response to the stimulus with no negative implications like stress or suffering. In a review of the literature, Mellor [60] argues that foetuses spend a lot of time in sleep, and this would preserve them from pain. But two pieces of evidence contrast with this conclusion: foetal sleep is not continuous and the foetus can be awakened [51, 61, 62]. Term foetuses spend 9% of the day awake [45]. F3 (calm wakefulness) and F4 (active wakefulness) states are present in 21% of the day in full term foetuses [63]; several reports show that foetal sheep spend a significant amount of time in wake, during late gestation [64, 65, 66, 67, 68, 69, 70, 71, 72]. Even circadian rhythms are present in foetus after their midgestation, in particular with regard to heartbeat and locomotor activity [73]; also the EEG shows a wake pattern after the 30th GAW [74]. External stimulations [75] can wake up even sleeping foetuses [76, 77]. It is common knowledge that childbirth, and the manoeuvres that caregivers perform on them at birth, cause awakening in all viable foetuses.
According to some authors [60, 78], progesterone, prostaglandins and adenosine present in the amniotic fluid cause foetal sedation and therefore analgesia. Nonetheless, this raises some perplexities. The first is the confusion between the terms “sedation” and analgesia”: the mere fact that a patient is sedated does not imply that they are not feeling pain. Sedation is a state of diminished awareness, not of analgesia. But there is something more: these neuromodulators, should be at foetal concentrations greater than in mothers’ blood to have an actually anaesthetic effect: if mothers’ blood values are not enough to be analgesic, a higher level is mandatorily necessary to provoke analgesia in the foetus. But human foetal blood adenosine levels in the third trimester of gestation are only moderately higher than those in mothers’ blood (0.58 microM/L [79] vs 0.41 microM/L [80] respectively), though in sheep maternal levels “2–4 folds greater” than in foetal blood are reported [81]. Furthermore, Yoneyama et al. [82] showed that blood adenosine levels are higher in pre-eclamptic mothers than in foetuses, though this does not cause analgesia in the mothers. Progesterone is higher in the foetal blood than in maternal blood across pregnancy, but this difference almost disappears at the end of pregnancy: in fact, while foetal progesterone remains constant, maternal progesterone increases with gestational age [83]. We should point out that, though higher than usual, these values are utterly far from being analgesic in mothers, and consequently in foetuses. Prostaglandin D2 has sleep-promoting properties, but the only studies that assess this property are performed by intracerebral infusion [84].
Several studies assess the changes in stress hormone production due to pain or stress in foetal animals. The catecholamine system has been studied to investigate if pain provokes an increase of adrenaline/noradrenaline in a similar way as in adults. In effect, stress can provoke an increase of these hormones in animal foetuses during the development period correspondent to the second half of human gestation [85, 86]; studies performed on monkeys show that this production is accompanied by long-lasting adverse changes in motor, social and cognitive behaviour [87, 88, 89, 90]. Other studies were performed in human foetuses. Stress hormones soar during a blood transfusion performed in human foetuses through their hepatic vein; this increase is not present if the transfusion is performed through the umbilical vein that has no nociceptors [91] and it is also absent if the foetus was previously given opioids [92]. Some authors argue that this does not necessarily mean a response to pain, but just a response to a physical stimulus with no emotional implication. This would mean that anesthetised adults have a similar increase in stress hormones if they receive a painful stimulus during the influence of anaesthesia. On the contrary, adult anesthetized patients show no increases in stress hormones during surgery [93, 94]. Marana et al. [95] performed a review of the literature showing an increase of stress hormones in patients undergoing surgery, but most of the articles cited assessed their measures in the hours or days after surgery (i.e., including post-surgery stress), but not during surgery [96, 97, 98]. Only one article [99], showed an increase in adrenaline and noradrenaline during anaesthesia in adults, but only in the group of patients who underwent a special technique (carbon dioxide insufflation) for laparoscopic resection of ovarian tumours; another paper questioned the effectiveness of low doses of anaesthetics to blunt adrenaline increase in adults, but agrees that high doses are effective in doing it [100]. Painful stimulation causes an increase in catecholamines in anesthetized brain-dead patients, although a missed increase in the control group decreases the strength of this study [101]; moreover, brain death itself can have been the cause of this increase [102].
In the third trimester of gestation, foetuses show signs of some kind of consciousness. To assess this, researchers introduced an auditory paradigm for the investigation of memory traces over different time scales, the so called “local-global” paradigm [103]. The local-global paradigm consists of a sequence of tones which can either contain only identical tones or end with a different tone. In practice, authors studied if foetuses responded with a startle to a new sound inserted in a series of equal sounds (first order regularity violation) or if they responded with a startle when they perceived a new sound sequence after a series of equal sound sequences (second order regularity violation). The detection of an error can be assessed by the event-related P300 component in the electroencephalogram, an endogenous potential due to a person’s reaction to the stimulus. In particular, P300 is considered to be developed in mental processes involving the evaluation and categorization of a stimulus. The authors concluded that the foetal assessment of second-order regularity violation is a sign for conscious mental processing [104]. Another sign of likely foetal consciousness is the presence of a rudimental brain Default Mode Network in the last trimester of pregnancy, deputed to consciousness and to the development of empathy and theory of the mind [27, 105]. The Default Mode Network involves several brain areas, as the hippocampus, the parietal lobe, and the angular circumvolution [28]. These data show that the foetus has some kind of consciousness, and this makes its perception of pain to some extent similar to that of an adult.
Two premises should be made to deal with the use of anaesthetics for prenatal surgery.
Attention should be paid to the administration of maternal anaesthetics during the pregnancy [106], due to the risks highlighted in 2016 by the Food and Drug Administration, which suggested not to exceed the 3-hour exposure to anaesthetics for children under the age of 3 years [107]. This group potentially includes the foetal population exposed to anaesthetics during maternal anaesthesia. It is supposed that GABA agonist (propofol, thiopental, isoflurane) and NMDA antagonist drugs (ketamine, nitrous oxide, tramadol) may exert their toxic effect on the developing brain by an action on GABA and glutamate receptors [108, 109]. The effects of isoflurane on the developing brain were evaluated using foetal sheep models [108]; pregnant ewes were exposed to single or repeated doses of isoflurane: no significant neuroapoptosis was observed in the single exposure groups, but repeated isoflurane exposure resulted in increased neuroapoptosis of the frontal cortex [108]. Prolonged exposure to isoflurane produced neuroapoptosis in fetal brain [110] as well as in the neonatal brain [111]. The negative effect on foetal heart has been reduced by administration of supplemental intravenous anesthesia (SIVA) with propofol and remifentanil infusions prior to uterine incision [112] or of just remifentanil [113] to the mother: both lower the required amount of desflurane for anaesthesia. Sevoflurane can damage the fetal frontal cortex in mice [114]. Some anaesthetic drugs have a protective effect on the brain and can be useful in preventing the negative effects of anaesthesia in foetuses. Dexmedetomidine has no interaction with GABA or NMDA receptors, and reduces the doses of volatile anaesthetics while providing anxiolysis, hypnosis and analgesia. Bypassing both GABAergic and glutaminergic systems makes dexmedetomidine, an alfa2-agonist, a possible neuroprotector of the foetal brain from isoflurane. This effect has been showed in ovine and foetal rat models [115]. Xenon, a noble gas, is another potentially neuroprotector [116]. Even propofol can give some neuroprotection when used in addition to isoflurane alone [112]; nonetheless, propofol detrimental effects on the developing brain have been extensively described [117].
The extraction index represents the amount of medication that is removed from the foetal circulation by the placenta and does not arrive to the fetus. In the case of maternal anaesthetics, it varies and can be high; in this case, even though the anaesthetic given to the mother attains good values in her blood, it is not sufficiently high in foetal blood [118, 119].
The values available for the foetal/maternal total concentration ratios were approximately 0.3 for bupivacaine and etidocaine, 0.5 for lidocaine, 0.7 for mepivacaine, and 1 for prilocaine [120]. The low ratios of bupivacaine and etidocaine result from extensive binding (90%) of these drugs to maternal alpha1-acid glycoprotein which exceeds corresponding foetal protein binding (50%). The passage of drugs through the placenta depends on various factors, which will affect the concentration of the drug in the foetal blood [120]. These factors are fat solubility, molecular weight and ionic charge of the drug. The highly ionized molecules, the liposoluble ones and those with molecular weight lower than 500 Daltons easily cross the placenta [121]. The passage of the drug also depends on the acidity of the foetal blood, which will be more acidic the more it will favor the passage of non-ionized molecules [121]. For example, lidocaine is a weak base that increases in foetal blood in case of foetal acidosis. Placental transfer of muscle relaxants is very low, with a foetal concentration ~10–20% of the maternal plasma concentration [122]. The extraction indices of halothane, isoflurane and nitrous oxide are 0.71/0.87, 0.71 and 0.7/0.8, respectively [122] and sevoflurane 0.38 [123]; that of propofol ranges 0.5/0.8 [124]; maternal blood concentrations of propofol are 14 times higher than those of foetal blood after 5 min after infusion and twice as high after 180 min [124]. Morphine and fentanyl extraction rate range 0.6–1 and from 0.5–0.9 respectively [123, 125, 126]. Table 1 summarizes the main evidences.
Drug | Extraction index (foetal/maternal ratio) |
Propofol | 0.7–1.3 |
Ketamine | 1.26 |
Thiopenthal | 0.4–1.1 |
Morphine | 0.61–1 |
Fentanyl | 0.5–0.9 |
Oxycodone | 1 |
Remifentanil | 0.29–0.88 |
Legend: see the text for details. From Ref#123, adapted. |
We should also discuss if the foetus requires a direct or an indirect analgesic treatment.
Several authors endorse a sole analgesia given to the mother to anaesthetise
both mother and foetus [127], in particular in the case of laparoscopic surgery
[128]. Others suggest an additional foetal direct analgesia should be
administered during surgery [129] using opioids. Fentanyl provided
intramuscularly to the fetus (10
Last, foetal activity during surgery is worth monitoring, with regard to general suffering and to pain.
The need of foetal monitoring is a tenet in this field. We should be aware if the foetus is undergoing any hypoxemic or painful distress during this kind of surgery. To this aim, we need a continuous monitoring of the foetal heart rate and foetal heart variability, being the latter a sign of sympathetic unbalance [137]. Nonetheless, a change of foetal heart rate variability is not always an indicator of foetal distress but may simply be a sign of expected anaesthetic effects on the foetal autonomic nervous system [138]. Cardiotocography, fetal echocardiography or ultrasound assessment of umbilical or middle cerebral artery blood flow can be monitored for fetal well-being, but interpretation can be difficult [139]. Fetal blood pressure monitoring is not yet feasible [139] Slowing of the fetal heart can be a sign of fetal hypoxaemia and acidosis, but can also be related to a decrease in temperature, maternal respiratory acidosis, or to the administration of drugs [140]. Some authors have tried to create a fetal pain assessment tool, to be used during prenatal surgery [141]. The current fetal monitoring during prenatal surgery is based on echocardiographic monitoring of fetal heart rate and contractility. If any concerns are indicated by the fetal echocardiographic evaluation, more advanced ultrasound monitoring of umbilical and middle cerebral artery are obtained.
The arguments in favour of foetal pain and the administration of foetal analgesics during open surgery overweigh those that are against. This should be taken into consideration by those who perform this special type of surgery. Signs of pain in human foetuses are evident from the 20th to 22nd week of gestation and the foetus should receive the same analgesic care during the surgery that a premature baby at an identical postconceptional age receives. Today the debate focuses on the type of anaesthetic to be provided to the foetus and whether administering anaesthetics to the mother is sufficient to guarantee foetal anaesthesia. In this review we have brought valid elements for those who have to perform this type of surgery.
Not applicable.
Not applicable.
This research received no external funding.
The author declares no conflict of interest.