“Few things a doctor does are more important than relieving pain” (1). The clinical use of opioids to treat pain has changed little since the days of Osler, except for the development of many synthetic compounds such as methadone, fentanyl, and alfentanil (1a). Morphine remains the mainstay of clinical treatment in many settings and the prototype opioid whose effects, both desired and undesired, are familiar to clinicians. They recognize that achieving the desired degree of analgesia often leads to tolerance requiring dose escalation accompanied by hypoventilation and decreased gastrointestinal motility. The symptoms of opioid withdrawal in newborns of mothers addicted to heroin, morphine, or methadone are familiar to pediatricians. The clinical problems associated with opioid therapy in pediatrics were reviewed recently by Suresh and Anand (2). Recent investigations have revealed many of the mechanisms involved in these processes, which have lead to potential therapeutic options to reduce tolerance and enhance analgesia.

Several reviews have been published recently that describe and diagram intracellular events associated with opioid analgesia and tolerance (2–6). Discussions of the differences between the proposed mechanisms in these reviews show that our understanding of these processes is still incomplete (7, 8). Some of the disagreements about the mechanisms of opioid analgesia and tolerance reflect the models used in individual studies or differences in study conditions (7, 9). Analgesia and tolerance involve integrated systems within the body that may exhibit autoregulatory changes extending well beyond the events in single cells (9). As Trujillo recounted from Umberto Eco, “for every complex problem, there's a simple solution, and it's wrong (7)!” There are no simple answers to explain the interplay between analgesia and tolerance associated with opioid therapy.

Although specific authors emphasize different portions of the processes involved in opioid-induced analgesia and tolerance, many of these intracellular events are generally accepted. These provide a useful framework to explain the effectiveness of specific therapeutic interventions to reverse tolerance and enhance analgesia.

Mu (μ), kappa (κ), and delta (δ) opioid receptors that are widely distributed in the brain, spinal cord, peripheral nociceptors, and other tissues belong to the steroid receptor family (2). These receptors contain seven transmembrane regions in which the extracellular regions provide receptor specificity while the intracellular portions link to inhibitory G proteins, Gi and Go, as well as to the stimulatory G protein, Gs (2–5). The signal from opioid receptor activation is transduced through ion channels for potassium, calcium, and enzyme systems in the cytosol and cell membrane (protein kinase C, adenylate cyclase, phospholipase A2, nitric oxide synthase, and possibly metabotropic glutamate receptors). Analgesia occurs as intracellular K+ increases, Ca++ decreases, and cyclic AMP (cAMP) decreases (2). These changes reduce the action potential duration and neurotransmitter release. Phospholipase A2 stimulation releases arachidonic acid that activates the lipoxygenase pathway to stimulate the inward K+ channel and raise intracellular K+. This provides a pathway for synergism between nonsteroidal anti-inflammatory drugs and opioids (2).

Tolerance to opioids may develop through multiple potential mechanisms that counter the changes associated with analgesia. Many of the changes associated with tolerance involve the excitatory amino acid, aspartate, pathways, and the n-methyl-d-aspartate (NMDA) receptor. Crain and Shen (3) outline a mechanism for shifting opioid receptor activation toward tolerance through GM1 ganglioside-related coupling of the opioid receptor output to the Gs proteins. This activates phosphokinase A, increases intracellular cAMP and Ca++, reduces intracellular K+, lengthens the action potential duration, and increases neurotransmitter release leading to tolerance and increased nociception.

Protein kinases A, C, and G protein-related kinases are central to the balance between tolerance and analgesia. Activation of the NMDA receptor through phosphorylation by protein kinases A and C opens the channel to increase intracellular Ca++. Feedback pathways that desensitize the Gi-Go inhibitor proteins also involve protein kinases A and C along with G protein-related kinases.

A recent review postulated that the opioid receptors are closely linked to three groups of postsynaptic metabotropic glutamate receptors (mGluRs) that in turn link to G proteins and phosphatidylinositol (4). An accompanying discussion points out that the mGluRs are often presynaptic autoreceptors that decrease glutamate release when stimulated, yet opioid receptor-mediated events are generally postsynaptic (7). Fundytus et al. (4) point out that intracellular increases in cAMP after acute opioid treatment modulate during chronic treatment. This can be explained by phosphatidylinositol-mediated desensitization of the NMDA and μ-opioid receptors through phosphorylation (4). Such differences between chronic and acute events further complicate the interpretation of single measurements of cAMP.

Tolerance and abstinence remain difficult clinical problems. All infants and young children treated with fentanyl infusions for 9 d or to a total dose of 2.5 mg/kg showed an abstinence syndrome when fentanyl was stopped (10). Increased understanding of the steps in these mechanisms of tolerance have been applied clinically and experimentally. The calcium channel antagonist, nimodipine, has been used to reduce opioid tolerance in patients with cancer (11). Opioid tolerance has been reduced also by treatment with NMDA receptor antagonists, such as ketamine (12), LY235959 (13), and loperamide (14).

Compared with adults, the developing infant may respond to opioids quite differently with potentially slower metabolism and excretion, greater brain penetration through an incomplete blood brain barrier, and differences in receptor binding and coupling. Few investigations of opioid tolerance have been conducted in developing animals. Choe and Smith (15) report an evaluation of opioid tolerance to analgesia as well as sedation in immature, 17-d-old rats implanted with fentanyl infusion pumps to deliver 60 μg/kg/h infusions of fentanyl for 72 h. Analgesia was studied with the tail-flick test in which a controlled heat source is applied to the rat's tail for up to several seconds. The interval from application of heat to moving the tail tests the degree of analgesia produced by increasing doses of fentanyl. Sedation was measured with two tests; the time to withdraw the rat's paws back from over the edge of a platform and the time to right itself after placing it supine. When tolerance develops, a greater dose of fentanyl is required to produce the same effect, which shifts the dose-response curve to the right. The immature rats, like adults, developed tolerance to analgesia after the 72-h infusion of fentanyl in high doses. Unlike adults, the 17-d-old rats surprisingly developed tolerance also to sedation. This sedative tolerance extended also to a new opioid, etorphine that is 5 000–10 000 times as potent as morphine, so increased receptor activity did not overcome the tolerance. Choe and Smith also investigated cross-tolerance to the sedative effects of the benzodiazepine, midazolam, in the morphine tolerant rats. In contrast to older rats that developed cross-tolerance to sedation between morphine and diazepam (16), they found no cross-tolerance in sedative effects between fentanyl and midazolam. Although the dose-response curves for sedation by midazolam in the fentanyl-infused rats were shifted to the right, the change was not significant. These differences between immature and adult rats' tolerance to opioids may be related to developmental differences in nervous system function, such as myelination or immaturity of the blood brain barrier. If similar tolerance to sedation accompanies tolerance to analgesia in infants, additional treatment beyond fentanyl may be required to achieve sedation.

Providing appropriate analgesia to pediatric patients, especially preverbal neonates and infants who cannot identify or quantify their pain, is a therapeutic challenge. To most parents and many pediatricians, it is intuitive that painful stimuli for adults will be painful for infants. Yet, pediatric patients, especially neonates, have undergone painful, surgical procedures without the benefit of the same analgesic treatment afforded adults (17). Emergent care sometimes prevents timely treatment, but many times we simply don't think of the need to provide adequate analgesics.

As pediatricians recognize pain in their patients as a legitimate therapeutic need and attempt to provide adequate analgesia, they are confronted with many challenges from the identification and quantification of pain to its treatment. Many unanswered questions about the basic pharmacology of narcotic analgesics persist and are confounded for pediatricians by the developmental changes in their patients. Does tolerance to sedation accompany tolerance to analgesia in neonates, while it does not in adults? Is tolerance to sedation unique to certain opioids? Is the mechanism for the development of tolerance the same in neonates and adolescents, and how can we reduce it? Answers to these questions and careful translation of observations like those of Choe and Smith to the bedside will both improve clinical care and provide important clues about the pharmacologic effects of opioids in our developing patients.