SOME scientific discoveries trigger a de rigueur consideration of physiologic principles (and our philosophical positions), often catalyzing the need for major changes in clinical practice. The elegant studies reported by Fisk et al. 1in this issue of Anesthesiology provide one such example by challenging the scientific precepts that have traditionally driven the clinical approach to fetal and perinatal medicine. This research group has taken advantage of a unique clinical situation to determine whether the endocrine stress responses associated with intrauterine exchange transfusions can be alleviated by fentanyl in the human fetus.
In specialized centers, hemolysis resulting from Rh-isoimmunization can be treated by exchange transfusions, performed on 2–5 occasions by cannulating either the fetal intrahepatic vein (IHV technique) or the umbilical vein at the insertion of placental cord (PCI technique). Transfusions via the IHV technique require the insertion of a needle through the abdominal wall, the peritoneal reflection, and the hepatic capsule, thus eliciting endocrine stress responses in the fetus (cortisol, β endorphin), 2in contrast to the PCI technique, which does not transgress any tissues with sensory innervation and does not seem to elicit any stress responses.
Although it would have been ideal to randomize all patients prospectively to undergo PCI or IHV procedures with or without fentanyl in a crossover design, this was precluded by technical considerations as well as by the fact that not all fetuses require multiple intrauterine transfusions. Therefore, the investigators performed a “paired longitudinal analysis” to compare the IHV-stimulated stress responses with or without fentanyl and an “unpaired cross-sectional analysis” to determine whether the fetal stress responses after IHV transfusion with fentanyl analgesia were comparable to those after the PCI technique. Their findings indicate that fentanyl attenuated the fetal β-endorphin stress responses, but the cortisol responses were not statistically different. One reason for this discrepancy could be that the study had insufficient power to examine a fentanyl effect on the more modest cortisol responses stimulated by IHV transfusions. Steroid biosynthesis in fetal adrenal cortex is immature, resulting in the secretion of precursor adrenocortical hormones in response to pain or stress, 3which were not measured. Activation of the hypothalamic–pituitary–adrenal axis from the release of corticotropin-releasing factor and related neuropeptides (e.g., arginine vasopressin, norepinephrine) also activates the sympathetic nervous system via the posterior hypothalamic nuclei. 4Perhaps Fisk et al. 1could have measured other endocrine markers (e.g. , adrenocorticotropin hormone, catecholamines, or the precursor steroid hormones) to better characterize the fetal responses to IHV transfusions.
The hormonal stress responses seen by Fisk et al. 1do not necessarily indicate fetal pain perception, nor does a fentanyl dose (12.5 μg/kg of estimated fetal weight) administered after IHV cannulation equate with fetal analgesia or anesthesia. These authors have adequately discussed the technical constraints, limitations in study design, and alternative explanations for their results. Nevertheless, even to the skeptics, these data provide convincing evidence for pain-induced stress responses in fetuses between 20 and 35 weeks of gestation, confirming previous work by the same investigators 2,5–8and preliminary findings from others. 9For example, the pulsatility index of the middle cerebral artery decreased within 70 s after painful stimulation in fetuses from as early as 16 weeks of gestation. 7Such robust physiologic responses would be unlikely if human fetuses were impervious to the pain induced by IHV needling.
Despite the possibility of direct cardiovascular or hormonal effects, 10–12the most prominent effect of intravenous fentanyl is analgesia and sedation. Other physiologic changes in the fetus may be a consequence of its analgesic effect. 13The responses of fetuses given fentanyl in this study were comparable to the physiologic and behavioral responses of preterm neonates receiving fentanyl for analgesia and sedation, 3,14despite a lower gestational age (22–32 weeks). Therefore, for some clinicians at least, these data may indicate that the human fetus is capable of responding to pain, which can be treated by opioid analgesia. To be convinced, other clinicians may need to see functional magnetic resonance images or direct neurophysiologic recordings from the somatosensory cortex (or the dorsal horn, thalamus, or other brain areas) from fetuses undergoing invasive procedures with or without analgesia.
The clinical and philosophical importance of these findings hinges on the authors’ assertion that a functional pain system develops in the human fetus by the third trimester of pregnancy. Even if their nociceptive pathways and reflexes are physiologically active, do humans consciously experience pain from intrauterine needling? Does consciousness occur at birth or does it exist in utero ?
Fetal behavior in utero must be differentiated from that of premature infants because of the possibility that the process of birth and the demands of independent survival may trigger the expression of consciousness. Perhaps the widespread and abundant expression of c-fos and other genes immediately after birth 15reflect the neuronal correlates of consciousness “developing” at birth. Evidence from the studies of postnatal behavior in preterm infants, which show multiple parallels with the behavior and capabilities of term infants, could be used to support the hypothesis that consciousness develops at the moment of birth. 16However, the question of fetal consciousness is fraught with intense controversy. The British Commission of Inquiry into Fetal Sentience 17declared that fetuses may be conscious from 6 weeks of gestation, whereas the Royal College of Obstetrics and Gynaecology 18countered that fetuses cannot be considered sentient before 26 weeks of gestation. Hormonal or circulatory responses do not vouchsafe conscious pain perception, although their absence would be more likely if sensory stimuli from these invasive procedures were not reaching the thalamus and hypothalamus.
Afferent inputs can alter the activity of neurons in the neocortical alange by 20 weeks of gestation, when thalamocortical and cholinergic afferents form synapses with the upper subplate neurons, 19whereas noradrenergic and dopaminergic fibers start to penetrate the subplate zone by 13 weeks of gestation and reach the cortical plate by 16 weeks. 20Thalamocortical axons penetrate the primary somatosensory cortex by 24 weeks of gestation, 21providing the final anatomic link for the developing somatosensory system. Therefore, somatosensory evoked potentials were recorded from the sensory cortex of 25-week preterm neonates. 22From approximately 20 weeks of gestation, electroencephalo-graphic recordings and ultrasound studies can differentiate sleep states and wakefulness, 23,24as well as responses to touch 25and sound. 26Experimental paradigms investigating the prenatal acquisition of memories in the third trimester of pregnancy further support the concept of fetal consciousness. 27To us, all these lines of evidence suggest that fetal consciousness develops from about 20–22 weeks of gestation.
Accumulating data may confirm or refute these tentative conclusions. However, we believe that current practice should incorporate the use of some form of analgesia or anesthesia for human fetuses subjected to surgical or invasive procedures. Direct administration of fetal anesthesia is the exception rather than the norm; most commonly, the mother is given systemic analgesia or general anesthesia for the procedure. 28A few specialized centers currently perform numerous surgical procedures for the correction of anatomic malformations (e.g. , those leading to hydronephrosis, hydrocephalus), and the complexity, range, and numbers of these procedures seem to be increasing. 29As more and more anesthesiologists are called on to deliver anesthesia and monitor women undergoing these procedures, the needs of the fetal patient should also be kept in mind. 30–32
Major fetal surgery or repetitive invasive procedures performed without consideration for the analgesic and anesthetic requirements of the fetus 33may or may not have the same long-term consequences associated with prolonged or repetitive neonatal pain. These have been the focus of intense inquiry recently. 34–38Because of the epochal developmental changes occurring in the immature brain during the third trimester of pregnancy, exposure of the unanesthetized fetus to surgery or invasive procedures may have an increased potential for long-term neurodevelopmental consequences. 39Neonatal rats treated with opioids require much higher doses for subsequent clinical effects, 40–42whereas the long-term neurodevelopmental consequences of prolonged opioid exposure in ex-preterm neonates seem to be relatively benign. 43
The limited evidence available suggests that fetal analgesia, provided for short periods with judicious doses of opioids, may have relatively few long-term detrimental effects and should be given during invasive in utero procedures. With this line of investigations, Fisk et al. 1have opened the door for the development of an entirely new field of fetal anesthesia, requiring the development of newer clinical skills, innovative anesthetic techniques, and perhaps novel ways of examining immediate and long-term clinical outcomes.