Research programs involving perinatal (fetal and neonatal) animals offer insight into the development of the brain and central nervous system and the age-dependent effects of genes, toxicants, and the environment. Such research has shed light on autism, learning disabilities, and fetal alcohol syndrome. Experiments involving perinatal animals use the same techniques discussed in other chapters and therefore pose the same animal care and use concerns as experiments involving adult animals. However, three issues influence how the care and use concerns are addressed in perinatal studies (NIH, 1991). First, perinatal physiology can be radically different from adult physiology, and it changes throughout early development, affecting such aspects of animal use as appropriate euthanasia and analgesia. Second, perinatal studies often entail the use of animals at a variety of developmental stages, which differ physiologically, as in the permeability of the blood-brain barrier (Saunders et al., 2000); this may necessitate that animals of different ages be cared for differently even if they are used in the same experiment. Third, when fetal studies are proposed, the welfare of both the mother and the fetus must be considered.
DEVELOPMENT OF PAIN PERCEPTION
The development of neural systems essential for pain perception has been studied most extensively in the rat, as has the early development of pain-related behaviors. Although there is no definitive evidence that prenatal animals can perceive pain, reflexive behavior in fetal animals sometimes correlates with behavior exhibited by adult animals in response to pain stimuli. It is not known
when developing animals begin to perceive pain. Reflexive withdrawal from noxious stimulation is observed in rodent embryos starting in late gestation, for example, on embryonic day 17 (E17) in the rat fetus (Narayanan et al., 1971). Human fetuses develop stress hormonal and circulatory changes in response to noxious stimuli by 18–20 weeks of gestation; similarly, fetal lambs and rhesus monkeys demonstrate changes in the pituitary-adrenal axis after application of stressors at the late gestation ages of 125 days and 133 days, respectively (Rose et al., 1978; Smith et al., 2000). Behavioral responses to injection of an irritating substance (formalin) into the paw can be seen in rat fetuses as early as E19, and the response correlates with expression of the c-fos protein (an indication of neuronal activation) in the spinal cord by E20 (Yi and Barr, 1997). By birth, neural substrates for perception of noxious stimulation are present in the periphery and spinal cord of the rat pup, although sensory systems are immature and undergo substantial change during the first few weeks after birth. Many neurotransmitters and receptors in pain pathways appear early in development, but their expression may vary—in either direction—during the neonatal period and may take weeks to achieve adult levels. One could argue that the physiologic response to noxious stimuli suggests a correlation with sensitivity to pain (Mahieu-Caputo et al., 2000; Smith et al., 2000). In theory, cortical recognition of pain in a human fetus should occur in the 26th week of gestation with development of thalamocortical connections (Vanhatalo and van Nieuwenhuizen, 2000).
Rat pups show behavioral arousal and withdrawal responses to noxious thermal and mechanical stimuli as early as the first postnatal day (Barr et al., 1992; Blass et al., 1993; Fanselow and Cramer, 1988; Fujinaga et al., 2000). In addition to behavioral responses to injection of irritating chemicals (Abbott and Guy, 1995; McLaughlin et al., 1990), rat pups as young as 3 days show allodynia and hyperalgesia in response to experimentally induced inflammation (Marsh et al., 1999). The behavioral responses of young pups to noxious stimulation are mostly generalized, whole-body responses, such as wriggling, although more localized withdrawal responses are also seen. As pups mature, their responses become more organized and localized and more typical of adult responses.
Dampening of behavioral responses to noxious stimuli, particularly when such opioid drugs as morphine are administered, is also seen within the first few days after birth (McLaughlin et al., 1990; Fanselow and Cramer, 1988) and the sedating effects of such drugs as pentobarbital can be distinguished from the analgesic effect of morphine as early as postnatal day 1 (P1) in the rat pup (Abbott and Guy, 1995). Mature responses to analgesics are seen around the age of 3 weeks in the rat, coinciding with the maturation of supraspinal descending inhibitory processes. Little information is available regarding neonatal precocious mammal responses to analgesics during postnatal development; however, most neonatal animals develop physiologic responses that are consistent with adult responses by the age of 6–8 weeks. In addition, many physiologic differences between neonatal and adult animals—such as the neonate’s greater perme-
ability of the blood-brain barrier, higher body water content, less mature hepatic microsomal enzyme systems, and lower albumin concentrations—affect the pharmacodynamics and pharmacokinetics of analgesic and anesthetic drugs (Thurmon et al., 1996).
In summary, although there is not enough evidence to determine whether neonatal animals perceive pain, some stimuli that are noxious to adult animals have been shown to trigger reflexive behavior in neonates, and this suggests that neonates would benefit from the administration of analgesics.
ANESTHESIA AND ANALGESIA
Available evidence suggests that the late-term fetus (E19–E20 in the rat) is responsive to noxious stimulation, as is the late-term fetal lamb and 26-week human fetus. Therefore, provision of anesthesia for potentially painful procedures is advised for late-term fetuses. For fetal manipulations in utero, anesthetics used to prevent pain in the mother are probably adequate to prevent pain in the fetus. Most drugs used for anesthesia in mammals—including barbiturates, ketamine, opioids, and inhalant anesthetics—readily cross the placenta. Therefore, the primary consideration should be adequate anesthesia, analgesia, and supportive care for the dam.
Anesthetic agents widely accepted for use in fetal surgical procedures include such inhalants as halothane, isoflurane, and desflurane (Abboud et al., 1995; Sabik et al., 1993). Balanced anesthesia with isoflurane and thiobarbiturates has been successfully used for late-term fetal pigs (Sims et al., 1997), whereas methoxyflurane and xylazine are associated with postnatal mortality in puppies delivered by caesarean section when those drugs were used for anesthesia in the dam (Moon et al., 2000).
Monitoring of anesthesia in fetal animals presents several challenges. Electrocardiographic monitoring can be used to easily monitor heart rhythm and electrical activity in fetuses of larger mammals, though bradycardia is a poor indicator of fetal distress. Pulse oximetry is noninvasive and effective for fetal lambs. It has a rapid response and is simple to use on the exposed fetus of larger mammals (Luks et al., 1998b). Direct monitoring of blood pressure and intravascular oximetry can quickly and accurately indicate fetal distress but are generally considered impractical because of their invasiveness.
In some studies, consideration of the potential effect of in utero drug exposure on physiologic and behavioral development of the animal may be appropriate (e.g., Belcheva et al., 1994; Niesink et al., 1999; Rodier et al., 1986). Non-opioid analgesics such as acetylsalicylic acid and acetaminophen, are potent inhibitors of prostaglandin synthesis, and their use in a fetus may result in unintended physiologic effects (Peterson, 1985). Prenatal administration of meperi-
dine or bupivacaine to primates may influence behavioral maturation (Golub, 1996). In higher mammals, such as nonhuman primates, appropriate postoperative analgesia for the dam is an important precaution in preventing premature labor after intrauterine surgery (Tame et al., 1999).
For experimental protocols that require the manipulation of late-term rodent fetuses after their removal from nonanesthetized mothers (such as a dam euthanized by decapitation or cervical dislocation), guidelines for anesthesia and analgesia in neonates should be followed.
Potentially painful experimental manipulations in neonatal rodents require the use of anesthesia or analgesia unless the IACUC has approved withholding anesthesia or analgesia for scientific reasons. The primary difficulty in using anesthesia or analgesia in the neonate is balancing its effectiveness and safety. Many anesthetics that can be used safely and effectively in adult rodents are not good choices for neonates; two examples are pentobarbital and ketamine, both of which tend to be ineffective at lower doses and fatal at higher doses (Danneman and Mandrell, 1997). In general, neonatal rodents are more sensitive to anesthetic and analgesic drugs than are adult animals, and such toxic effects as respiratory and cardiac depression are more serious problems in the youngest animals (e.g., Colman and Miller, 2001; Fortier et al., 2001; Greer et al., 1995; Prakash et al., 2002).
Most of the anesthetic agents used in juvenile and adult animals are safe and effective in larger neonatal mammals (Grandy and Dunlop, 1991; Thurmon et al., 1996). The choice of anesthetic agent used may depend on species, type and duration of procedure, and availability of specialized equipment needed (such as a gas anesthesia machine with a precision vaporizer). Most anesthetic regimens used in precocious and nonrodent neonatal mammals are standard veterinary procedures.
The choice of anesthetic method or agent should be based on the procedure, expertise of the researcher, the potential for hemorrhage, and the stability of the anesthetic plane. Overall, the best results of anesthesia in neonatal rodents have been achieved with inhalant anesthetics and hypothermia. Inhalants are a reasonable first choice for anesthesia of neonatal rodents. When inhalants cannot be used—for safety or practical reasons—hypothermia should be considered as a safe and effective alternative to injectable drugs. Hypothermia has been proved safe and effective as the sole method of anesthesia for altricial rodents (such as rats and mice) up to about the age of 7 days (Danneman and Mandrell, 1997; Phifer and Terry, 1986). However, it has the potential to be noxious, and rapid cooling of nonprotected flesh is painful (Wolf and Hardy, 1941). Rat pups recovering from hypothermia—but not pups recovering from general anesthesia—emit ultrasonic vocalizations even when placed with their mothers during the recovery
period (Hofer and Shair, 1992). The significance of the vocalizations is not clear, but they may indicate distress. To reduce possible unintended pain associated with cooling, the technique for inducing hypothermia should include partial insulation of the pup (for example, by wrapping in a latex blanket) (Danneman and Mandrell, 1997).
As with adult animals, assessing the effectiveness of anesthesia in neonates is important before beginning a potentially painful procedure. Adequately anesthetized rat pups will not respond to a light pinch of the foot or tail. Similarly, adequately anesthetized adult rats will not respond to a pinch of the toe or tail.
Opioid drugs provide effective analgesia against thermal, inflammatory, and mechanical pain in neonatal rodents as young as P1 (Barr, 1999; Barr et al., 1992; Helmstetter et al., 1988; Marsh et al., 1999; McLaughlin and Dewey, 1994) and should be considered for use whenever analgesia would be provided for an adult animal. Fentanyl is a recommended analgesic for neonatal dogs and humans because it has less of a respiratory depressant effect than morphine (Luks et al., 1998a).
Neonatal exposure to pain, especially when pain is an unintended outcome, may have developmental effects on the central and peripheral nervous systems and alter behavior and the threshold for pain in adulthood (Anand et al., 1999; Bhutta et al., 2001; Fitzgerald and Beggs, 2001).
SURGERY, POSTOPERATIVE MONITORING, CANNIBALISM, AND NEGLECT
Aside from the technical difficulties associated with using very small animals, surgical procedures involving neonatal rodents present such challenges as maternal neglect and cannibalism. As with adult rodents, pups should be kept warm, dry, and well hydrated postoperatively. They should be placed in a warm— not hot—environment until they have regained the ability to right themselves when placed on their backs or sides, after which they should be returned to their mothers. Some rodent mothers (particularly in some strains such as BALB/c mice) will reject or kill their pups under these circumstances. Some steps can be taken to reduce that problem. First, pups should be sufficiently recovered from anesthesia that they are able to right themselves and respond to stimulation. Smearing a pup with bedding and urine from littermates that remained with the mother can be helpful, as can placing the pup in the middle of the litter and allowing it to settle in for a minute or two before reintroducing the mother. Other methods that may work include masking olfactory cues by sprinkling baby powder on mother, pups, and bedding and smearing the pups and the mother’s nose with an aromatic agent, such as Vicks Vapo Rub®.
The following method is cumbersome, but it can greatly improve the rate of successful reunion of mouse pups with their mothers and might be considered when maternal neglect of pups is substantially inhibiting progress of a study:
When pups are removed for surgery, similarly aged pups can be taken from an outbred mouse (such as CD-1) and transferred to the mother whose pups were taken for surgery. Postoperatively, the surgically altered pups are then placed with the outbred mother for temporary fostering during the recovery period and left with her for a couple of hours or overnight. The litters are then switched so that each mother has her own pups back. Care must be taken to treat both experimental and control pups in the same way to avoid introducing experimental variability.
In any event, the mother’s behavior toward the pup should be observed closely for the first 10–15 minutes after the pup is returned to her and then every 10–15 minutes for the next couple of hours. At the first sign of aggression by the mother toward the pup, the pup should be removed. If other means of caring for the pup (such as fostering or hand rearing) are not available, the pup should be euthanized, as should pups that are not being cared for by their mother.
In higher mammals, neglect and cannibalism are uncommon postsurgical problems. However, the behavior of the dam should be closely monitored after return of the neonate to her.
IDENTIFICATION, TAGGING, TATTOOING, AND TOE CLIPPING
Two of the most common methods of identifying adult rodents, ear notching and ear tagging, are not useful for neonatal rodents, because they have small ears tightly placed against their heads. Temporary identification of hairless neonates can be achieved with nontoxic indelible markers (for example, Sharpie®). However, this marking rarely lasts for more than a day, because the mothers will lick the color off. More permanent identification can be achieved by marking the tail with a tattoo machine designed for this purpose; with practice, pups can be marked quickly and effectively. According to the Guide, “toe clipping [removal of the first bone of certain toes, corresponding to a predetermined numbering code], as a method of identification of small rodents, should be used only when no other individual identification method is feasible and should be performed only on altricial neonates” (p 46). Under some circumstances, that method of identification may be necessary, but it should be used only with IACUC approval based on appropriate justification in the animal-use protocol.
REGULATORY CONSIDERATIONS IN FETAL SURGERY
Many experimental fetal surgical procedures in higher mammals require special procedures or conditions, such as a second surgery for the injection of tracers or producing a lesion, or specialized equipment and facilities. Exposure of a fetus in utero constitutes a major operative procedure as defined by the AWRs and the Guide. In accordance with regulatory requirements for surgery, multiple survival surgical procedures must be justified scientifically by the neuroscientist
in the animal-use protocol and approved by the IACUC. In addition, traditional tracer injections, lesions, or recording may require that the surgical procedure be conducted outside facilities dedicated for aseptic surgery (such as in a laboratory setting). This represents a deviation from the Guide and the AWRs, so approval for such procedures rests with the IACUC. Performance standards and a team approach by the IACUC, the veterinarian, and the investigator can ensure that the spirit of the regulation is met and that veterinary care will not be compromised as a result of surgical procedures conducted under non-aseptic conditions (see “Asepsis and Physical Environment” in Chapter 3 and “Modified Surgical Settings” in Chapter 4).
Laboratory animals can be euthanized in three ways: hypoxia, depression of neural activity necessary for life function, and physical disruption of brain activity and destruction of neurons necessary for life (Balaban and Hampshire, 2001). However, the physiology of the perinatal animal renders some of the euthanasia methods used for adult animals inadequate and therefore inadvisable for perinatal animals (NRC, 1996).
In rodent fetuses that are less than E14, the lack of neural development prevents signs of fetal response to noxious stimuli, so euthanasia of the dam or removal of the fetus from the dam will result in the painless death of the fetus without a requirement for additional measures (NIH, 1997).
Inhalant agents, including inhalant anesthetics and CO2, that cause death by cerebral depression and/or hypoxia, must be used carefully for euthanasia of older fetuses or neonates. The comparatively hypoxic intrauterine environment renders these young animals much more tolerant of hypoxic conditions than adults (Singer, 1999), and euthanasia with an agent that causes death by hypoxia, such as CO2, may take 30 minutes or longer. Therefore, if these agents are used, personnel should be appropriately trained to use prolonged exposure times. Ideally, death should be verified by a secondary method such as decapitation or cervical dislocation.
Older fetuses and neonates can also be euthanized with chemical anesthetics, decapitation, or cervical dislocation. If chemical fixation of the whole fetus is necessary, the fetus should be properly anesthetized before fixation (NIH, 1997). In accordance with the report of the AVMA Panel on Euthanasia (2001), some physical methods of euthanasia, such as decapitation, require appropriate training, experience, and specific approval by the IACUC.