| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesia, Children’s Hospital, Boston, Massachusetts
Address correspondence and reprint requests to Charles Berde, MD, PhD, Department of Anesthesia, Children’s Hospital, 300 Longwood Ave., Farley 3 South, Boston, MA 02115.
Infants and children have been described as "therapeutic orphans" (1) because, historically, they have had limited support for inclusion in clinical trials of drug therapies, leading to a limited body of information regarding safe and effective prescribing practices. The vast majority of medications used for children do not have pediatric labeling, and in most cases, the package inserts or listings in the Physicians’ Desk Reference include phrases to the effect that "safety and effectiveness are not established for children ages 12 and under." The history of pediatric and perinatal therapeutics is replete with examples of drugs whose toxicities for infants and children were not anticipated by adult studies: congenital limb malformations caused by thalidomide, bilirubin encephalopathy caused by sulfonamides in neonates, "gray baby syndrome" from chloramphenicol, and bone and tooth deformities from tetracycline.
The process of introducing new drugs for human use begins with Phase 1 studies. These typically involve drug administration to consenting adults, either as healthy volunteers or as patients. In the case of healthy volunteers, there is little or no expectation of a specific therapeutic benefit for the participant, but there is the potential for adverse effects. As a dramatic example, consider the studies by Scott (2) and Scott et al. (3) conducted to determine surrogate measures of cardiac and central nervous system risks of bupivacaine and ropivacaine. The investigators administered incremental doses of these local anesthetics IV to adult volunteers until they observed initial signs of systemic toxicity, including symptoms of central nervous system depression and electrocardiographic changes. This study confirmed observations in animal models, indicating that ropivacaine may have comparatively less cardiotoxicity than bupivacaine. However, it would be practically and ethically impossible to perform this sort of study in infants and children. Instead, pediatric clinical trials generally involve the administration of drugs to patient populations for whom the drug has potential benefit for the participants. Initial dosing is often extrapolated from adult studies using formulas based on body weight or surface area. Ethical, financial, and logistical considerations often restrict the sample size of many pediatric studies. Thus, numerical estimates of the risk of adverse reactions in children may be very imprecise.
For these reasons, it is attractive to pursue studies of drug safety and efficacy in developing animals. Infant animal studies have several theoretical advantages:
The choice of species for developmental studies varies according to the questions of interest. For example, in studies of the fetal and transitional circulation relevant to pediatric cardiology and neonatology, fetal and newborn lambs and piglets have been especially important (6). Lambs and piglets have not commonly been used for studies of analgesia and nociceptive functions. Adult rats are extensively used in studies of nociceptive mechanisms and analgesic pharmacology, and as a result, the infant rat has recently been used for developmental studies in these areas. The advantages and potential pitfalls of the infant rat as a model are illustrated by considering three lines of research:
Development of Peripheral and Spinal Nociceptive Functions
Fitzgerald and her coworkers in London have conducted a series of studies in the past 15 years on the ontogeny of nociceptive functions in infant rats and infant humans. In many respects, these studies are the best examples of how to use and interpret a developing animal model. At birth, maturation of peripheral and spinal somatosensory functions in infant rats roughly parallels that of 24-week-old premature humans. The first seven days of postnatal life in rats correspond to the last 16 weeks of gestation up to term in humans. Rats are weaned at approximately 15 days of life and become sexually mature at approximately 21 days of life.
Fitzgerald et al.’s studies in rats and humans suggested that, in both species, infants were more reactive to certain noxious stimuli than adults. Infant and adult humans and animals, when presented with a sharp stimulus in an extremity, move that extremity toward the center of the body in a reflex limb movement known as the "flexion withdrawal reflex." Infant rats, when compared with adult rats, have lower thresholds for elicitation of flexion withdrawal reflexes (behavioral measurements). Spinal nociceptive neurons in infant rats, compared with adults, have wider cutaneous receptive fields, lowered thresholds for firing and prolonged after-discharges to noxious stimuli (electrophysiologic measurements). Low-threshold myelinated A alpha-beta fibers in infant rat synapse in the superficial lamina of the dorsal horn; these connections are not found in adult rats in the absence of nerve injury (electrophysiologic, neuroanatomic, and immunohistochemical measurements). Infant rats show an increased expression of the immediate early response gene c-fos in dorsal horn neurons by noninjurious as well as injurious stimuli (immunohistochemical measurements). Descending pain-inhibitory pathways in the dorsolateral funiculus mature comparately later than nociceptive transmission (neurosurgical interventions and electrophysiologic measurements); pain inhibition lags behind pain sensation. Infant rats show sensitization of nocifensive reflexes after tissue injury or inflammation to a greater degree than adult rats (14).
Thus, by using a combination of immunohistochemical, electrophysiologic, neurosurgical, and behavioral measures, a picture of nociceptive development emerges. The infant rat responds to noxious and nonnoxious stimuli with more brisk responses than adults, and they show less spatial and temporal precision in their pain responses compared with adults. Reflex responses in infant rats are subject to less supraspinal inhibition than in adults.
Two factors make Fitzgerald et al.’s observations especially compelling: 1) demonstration of anatomic similarities between rat and human somatosensory development by using human autopsy data and 2) a good correspondence between infant rat behavioral measurements and analogous behavioral measurements in human infants. Flexion withdrawal reflexes in premature human infants have lower thresholds than those in term infants for evoking a flexion withdrawal reflex. Premature infants undergoing invasive procedures (heelstick for blood sampling) develop secondary hyperalgesia that can be prevented by use of topical anesthesia. At earlier developmental stages, the withdrawal responses to noxious stimuli are more generalized, and bilateral responses can be evoked with a unilateral stimulus (15). With this correspondence between the behavioral measures in infant rats and humans, it becomes plausible to assume that some of the neurophysiologic mechanisms underlying these developmental changes in infant humans are also similar to mechanisms elucidated in the infant rat.
Ketamine and the Developing Brain
A report by Ikonomidou et al. (12) published in the journal Science in early 1999 has been of considerable concern to pediatric anesthesiologists. Professor Olney and coworkers in this group have, for many years, examined effects of both activation and inactivation of the NMDA subgroup of glutamate receptors on neuronal cell loss in development, with hypoxic-ischemic injury, and in psychiatric disorders. In the Science article (12), it was reported that several NMDA antagonists, including ketamine, increased programmed cell death, known as "apoptosis," in a wide spectrum of brain regions in rats before the age of seven days. Ketamine is widely used in pediatric anesthesia. It provides good hemodynamic stability when used for anesthetic induction, even among premature infants (16,17) and infants with congenital heart disease (18). It is thus important to address whether these infant rat studies give sufficient grounds for avoiding ketamine in neonates.
Programmed cell death has long been recognized as a normal aspect of neurologic development (19). Glutamate appears to be an essential neurotransmitter in early brain development. NMDA antagonists exert a myriad of actions. In other contexts, including studies of focal cerebral hypoxic-ischemic injury in infant rats, they have been reported to produce neuronal protection (20), rather than neuronal cell death. Nevertheless, a demonstration of 3- to 40-fold increases in cell death in many brain regions in the Ikonomidou et al. (12) study is a disturbing prospect for clinicians.
The clinical relevance of the Ikonomidou et al. (12) study may be called into question by considering how the duration of drug administration scales over developmental milestones. Although small size and short life span are convenient for conduct of experiments in many respects, they make this aspect of experimental design more problematic. Ikonomidou et al. (12) administered ketamine to seven-day-old rats using seven subcutaneous doses of 20 mg/kg for a nine-hour period. Nine hours is not a very unusual duration of general anesthesia in humans. However, when considered as a fraction of a critical period of time for neuronal migration and differentiation, nine hours in an infant rat’s life corresponds to an enormously longer period of time in an infant human. Depending on some assumptions about the choice of developmental milestones used for comparison, nine hours in an infant rat spans a period of neuronal development corresponding to roughly 8–40 days for infant humans. A recent study by Hayashi et al. (21), using somewhat different methodology, found that single 20-mg/kg doses of ketamine in seven-day-old rats produced no increase in brain cell death compared with control animals.
The small size of infant rats makes physiologic monitoring and stabilization technically challenging. Ikonomidou et al. (12) found that their infant rats appeared generally stable after their nine-hour anesthetics, but they could not exclude the possibility of physiologic instability, such as episodic hypoxemia, hypotension, or hypoglycemia occurring during these anesthetics. No mention is made in the article regarding respiratory or hemodynamic measurements, degree of sedation or analgesia, fluid or nutrient administration during these nine hours, or feeding behavior in the subsequent 24 hours before death. It is therefore an open question whether some of the neuronal cell death observed in these experiments represents drug action on the brain neurons per se or secondary consequences of hypoxemia, ischemia, or reduced substrate delivery during prolonged general anesthesia. Ikonomidou et al. (12) administered scopolamine, haloperidol, and nimodipine to control groups and did not find this degree of apoptosis. None of these latter drugs produce general anesthesia in most cases, and no control experiments were performed with a range of other general anesthetics. Thus, it remains to be determined whether prolonged general anesthesia per se would have similar effects in rats at this stage, with or without aggressive physiologic stabilization measures.
The clinical significance of this degree of apoptotic degeneration is unknown. Infant humans and animals have a remarkable capacity to recover after many forms of neurologic insult. What is not clear from the Ikonomidou et al. (12) experiment is whether there is an association between increased apoptotic neurons at this developmental stage and subsequent neurologic functioning. It would be instructive to perform behavioral and histologic experiments in adult rats treated with anesthetics in infancy.
Nitrous Oxide
The report by Fujinaga et al. (13) from Prof. Maze’s group in this issue of Anesthesia & Analgesia examines analgesic effects of nitrous oxide in infant rats. Prof. Maze and his colleagues have, for a number of years, made a great many important contributions to the study of molecular mechanisms of 
2-adrenergic agonists as analgesics, hypnotics, and general anesthetics. Their group previously demonstrated that nitrous oxide’s analgesic action is mediated, at least in part, via adrenergic pathways involving the locus ceruleus and descending tracts in the spinal cord. Other investigators have shown that spinal descending pain modulatory pathways, central and peripheral adrenergic receptors, and sympathetic nervous system reflexes are immature at birth in infant rats. It was thus natural to ask whether nitrous oxide would show a diminished analgesic effect in infant rats, compared with older rats.
In their current study, Fujinaga et al. (13) report an inability to show analgesic effects in rats before postnatal Day 21 using the tail-flick test. Nevertheless, as is evident from the work of Fitzgerald and colleagues, age-related changes in nociceptive thresholds and in tissue heat capacity may confound the interpretation of experimental results. Fujinaga et al. (13) attempted to compensate for these effects by reducing the intensity of the heat stimulus in younger rats to make their tail-flick responses comparable with those of older rats. However, a limitation of this study is its use of a single analgesic assay. Thus, for example, the results would have greater weight if nitrous oxide also failed to provide analgesia to brief noxious mechanical stimuli or inflammatory injury (22).
Twenty-one day-old rats can be regarded as roughly analogous to school-aged humans, depending on the choice of developmental parameter used for comparison. The authors thus conclude that nitrous oxide will not exhibit analgesic effects in infants and toddlers. Because nitrous oxide has some potential adverse effects, they conclude that it has virtually no favorable risk-benefit ratio in infants and, therefore, should not be used.
Although nitrous oxide certainly has a spectrum of risks and benefits, the authors’ conclusions appear inconsistent with results of clinical trials of nitrous oxide in infants and toddlers. Nitrous oxide reduces minimum alveolar concentration from volatile anesthetics at least as effectively in infants and toddlers as in adults, if not more effectively (23); the extrapolated minimum alveolar concentration for nitrous oxide in infants and toddlers was 109%, similar to values reported for adults. Among infants receiving large-dose fentanyl as an anesthetic for cardiac surgery, the addition of nitrous oxide improves the quality of anesthesia, as judged by suppression of autonomic and hormonal-metabolic responses (24). Nitrous oxide produces no increase in pulmonary vascular resistance in children with congenital heart disease (25), unlike adults with pulmonary hypertension.
The presumption of absent adrenergic responsiveness in newborn humans is also an oversimplification. Although adrenergic receptors and pathways and sympathetic reflexes are immature at birth, they are not totally absent. The addition of the 2 agonist clonidine to bupivacaine doubles the duration of spinal anesthesia in preterm infants undergoing inguinal hernia repairs (26). Among infants and toddlers receiving caudal blocks with local anesthetics, the addition of clonidine prolongs the duration of analgesia more than twofold compared with control groups receiving local anesthetics alone (27–29).
How much emphasis should be placed on the significance of infant animal studies in the absence of corroborating evidence in humans? Where results between infant rat and human studies clearly disagree, should we then completely ignore conclusions of infant animal studies? Researchers and clinicians should be cautious in changing clinical practice based on preliminary infant animal experiments in the absence of corroborating human data, but these infant animal experiments should cause them to question their practices and to pursue the corresponding human studies. The great value of infant animal studies is that they allow us to examine mechanisms beyond the limits imposed by the logistical and ethical constraints of human pediatric clinical trials.
References
This article has been cited by other articles:
|
|
 |
|
  S. M. Walker Pain in children: recent advances and ongoing challenges Br. J. Anaesth., July 1, 2008; 101(1): 101 - 110. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Que, P. J. Siddall, and M. J. Cousins Pain Management in a Patient With Intractable Spinal Cord Injury Pain: A Case Report and Literature Review Anesth. Analg., November 1, 2007; 105(5): 1462 - 1473. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Vutskits, E. Gascon, E. Tassonyi, and J. Z. Kiss Effect of Ketamine on Dendritic Arbor Development and Survival of Immature GABAergic Neurons In Vitro Toxicol. Sci., June 1, 2006; 91(2): 540 - 549. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. C. Hains and S. G. Waxman Activated microglia contribute to the maintenance of chronic pain after spinal cord injury. J. Neurosci., April 19, 2006; 26(16): 4308 - 4317. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
