Nociceptive Pain

Nociception refers to a neural response of nociceptors to a noxious stimulus.¹ Nociception is a normal nervous system function that serves as a warning of danger or credible threat. The nociceptive response comprises a nociceptor and three neuron chains, originating in the peripheral tissues and ending in the cerebral cortex. Nociception is unique among sensory nerves because nocicpetors must detect a wide range of stimuli, including physical and chemical signals, heat and cold, and acid and mechanical pressure. Nociceptors are located in every tissue of the body, originating from a group of neuronal bodies in the dorsal root ganglia (Figure 28-1). Nociception varies with location (peripheral versus central) and organ. Nociception in the skin occurs at the surface (i.e., is somatic), resulting in sharp, defined, localized, and limited pain. In contrast, internal, or visceral, nociception is generally diffuse, dull, and often referred.⁴ The action potential stimulated by a nociceptor crosses the synapse in the dorsal horn and stimulates the second order neurons in the gray matter of the spinal cord. There the signal is transmitted to the brain, where it is processed and interpreted. The transmission of pain from nociceptors is carried by either small, myelinated A delta fibers, smaller unmylelinated C fibers, or A a fibers. A delta fibers are fast, being responsible for sharp and acute pain, and transmit somatic and parietal pain. Because these receptors are discrete, animals can localize this pain. In contrast, C fibers are slow, transmitting dull, aching, burning, or throbbing pain that is difficult to localize. A fibers also are slow, transmitting stimuli associated with vibration, stinging, or tickling. However, because they are not able to discriminate between painful versus nonpainful stimuli, they are not pure nociceptor neurons.

Figure 28-1 Pain pathways and the role of peripheral nerve injury in the emergence of neuropathic pain. See text for the transmission and perception of pain (left). The right side demonstrates mechanisms whereby neuropathic pain might emerge. The lower left diagram delineates the impact of peripheral nerve injury on sensory neurons. Injury is followed by recruitment and proliferation of non-neuronal elements (including Schwann cells, mast cells, macrophages, and T cells). Cells release proinflammatory and other chemicals (TNFα tumor necrosis factor-alpha], IL[interleukin]-1, IL-6, CCL2, PGs (prostaglandins) and NGF [nerve growth factor]). Sensory abnormalities are initiated and maintained either locally at the site of damage or with retrograde transport to cell bodies in the dorsal root ganglion, where neuronal gene expression is influenced. Increased intracellular calcium results in activation of the p38 and MAPK/ERK pathway. Perpetuation of the response is maintained through recruitment of inflammatory cells in general and T-cells in particular. In the spinal cord (upper right), microglia surrounding primary afferent neurons contribute to self-propogation of neuropathic pain after activation by chemicals or the influence of other primary afferent neurons. Subsequent mediator release may presynaptically or postsynaptically modulate pain. Examples include but are not limited to TNF-α, IL-1, IL-6, nitric oxide (NO), ATP, and PGs and excitatory neurotransmitters (ENT). AMPA,amino-3-hydroxy-5-methyl-4-isoxazole propionic acid; NMDA, N-methyl-d-aspartate; MAP, mitogen-activated protein kinases.

Chemical mediators are important components of the nociceptor reflex and offer a target of pharmacologic modulation. These include but are not limited to adrenocorticotropic hormone, glucocorticoids, vasopressin, oxytocin, brain opioids, catecholamines, angiotensin II, endorphins/enkephalins, vasoactive intestinal peptides, substance P (centrally), eicosanoids (prostaglandins, leukotrienes), tissue kininogens (bradykinin), histamine, serotonin, potassium, and proteolytic enzymes.⁶ Several of these mediators are also associated with stress. Several neurotransmitters associated with nociception and transmission of peripheral pain also function in the dorsal horn, the first site of signal processing. Spinal neurotransmitters include but are not limited to peptides (e.g., substance P, calcitonin gene-related peptide), excitatory and inhibitory amino acids (aspartate, glutamate, gamma-aminobutyric acid [GABA]), nitric oxide, prostaglandins, adenosine-5’-triphosphate (ATP), endogenous opioids, monoamines (serotonin, norepinephrine), protons (acids), and neurotrophins. Opioid peptides are synthesized by interneurons in the superficial dorsal horn; they regulate further neurotransmitter release, probably through decreased calcium conduction. Glutamate, the primary excitatory nociceptor neurotransmitter, is released in response to calcium, following depolarization. Subsequent postsynaptic binding to N-methyl-d-aspartate (NMDA) receptors influences pain transmission, hyperalgesia, allodynia, and neurotoxicity, thus providing the rationale for combination analgesic therapy that includes NMDA-receptor antagonists.

The IASP definition of pain includes both a sensory and emotional component.² Indeed, of all the sensory systems, the nociceptive system is most able to elicit an arousal response in the brain. The emotional portion of pain occurs at the level of the limbic system (emotional), cortex (cognitive appraisal), and frontal lobes. Stimulation in the reticular formation results in emotional reaction to pain (anxiety, depression, suffering), whereas stimulation in the cerebral cortex leads to conscious perception and interpretation of pain. Transmission of central pain reflects a “gate control” phenomenon. The dorsal horn cells modulate the patterns of incoming information transmitted to the brain, which ultimately produces response and perception. The brain, in turn, can enhance or counter nociception. For example, when pain is sustained, released opioid peptides bind mu receptors in the brain and spinal cord, decreasing pain. Modulation regulates pain response, thus maintaining homeostasis.⁷ In addition to central transmission of pain, the first-order neuron can also synapse with neurons that cause a local reflex. The reflex can be myoneural or sympathetic in action (e.g., release of norepinephrine, smooth muscle spasm, vasoconstriction). Voluntary reflexes require conscious pain perception, whereas nociceptive reflexes do not.

Not surprisingly, a variety of mechanisms transduce nociceptive signals.⁸ At least two classes of channels are involved: One class detects noxious stimuli or products of tissue damage, and another sets the threshold necessary for a nociceptor action potential. Channels that detect noxious stimuli include transient receptor potential (TRP) channels, acid-sensing ion channels (ASICs), and the ATP-gated P2X receptor family. Of these, the TRP channels are emerging as the dominant channels transmitting nociceptor signals. The TRP channels are general receptors, able to sense multiple types of noxious stimuli. Their actions are complex, with channels having different roles in different tissues. Further, each channel is able to respond to more than one stimulus and with overlapping sensitivities. Four of the nine currently described nociceptive TRP channels respond to heat. The TRP receptors are regulated through a number of kinases (e.g., kinase C, kinase A, calcium/calmodulin–dependent kinase). In contrast to the general TRP channels, ASIC channels are specialists, able to detect only acidity, as might occur with bone cancer (high levels of acid leak into surrounding tissues), inflammation (local pH can drop to 5.5), and ischemia (resulting from metabolites such as lactic acid). Because extracellular ATP is a major product of inflammation, P2X channels are also drawing interest for their potential role in inflammatory pain, particularly visceral pain.⁷

The second type of channels that transmit nociceptor signals set thresholds and include newly discovered voltage-dependent sodium channels. Unlike other sodium channels, these channels are resistant to blockade by tetrodotoxin (TTX), which serves as a method of identifying and classifying the channels. Normally, these channels set the nociceptor threshold very high. However, in the presence of pathology, thresholds are decreased such that nociceptors respond to much weaker stimuli. Thus, unlike most sensory systems, channels responsible for nociception become more sensitive, rather than less sensitive, to stimuli. The TTX-resistant sodium channels tend to be redistributed after tissue injury.

Neuropathic Pain

Receptors

The existence of multiple opioid receptors was postulated on the basis of in vivo canine studies.³⁰ The pharmacologic effects result from interaction with one or more of three major opioid receptors named based according to the first letter of the first compound that bound to each receptor: mu (μ; morphine), kappa (κ; ketocyclazocine), and delta (δ). Receptors are similar in structure, expressing at least 40 homologies.The NC/OFQ receptor is a fourth member of the class whose similarity to opioid receptors is based on receptor homology. As with sigma receptors, they do not bind to classical opioid ligands. However, changes in as few as four amino acids would allow the nociceptin receptors to bind to classical opioids. Sigma receptors are no longer considered opioid receptors; they bind to and are activated by drugs completely unrelated to opioids. Opioid receptors also are chronologically classified, with OP-1 (δ) discovered first, followed by OP-2(к), and OP-3 (μ) (the most recently discovered). The Committee on Receptor Nomenclature and Drug Classification has indicated a final designation: MOP, DOP, and KOP, reflecting mu (μ), delta (δ), and kappa (κ) opioid receptors, respectively.²⁷ This latter classification will predominate in this chapter. Nociceptor receptors are classified as OP4 or NOP. Other opioid receptors (e.g., epsilon) may reflect splice variants or diamers.

Each of the opioid receptors is subtyped: μ-1, μ-2, к 1-3, and δ1 and δ-2, with differences in structure probably generated after translation. Receptors vary in several characteristics that complicate predictive responses. The μ-1 or MOP₁ receptors are described as very high affinity receptors that do not discriminate well between μ and δ. Dimerization among the receptors may play a role in their differential effects; thus far, both δ к and δμ diamers have been identified. Differential affinity of the heterodiamerized receptors compared to homodiamers for selected ligands (endogenous peptides or drugs) markedly alters the pharmacologic response. Further, binding sites of peptides are different from those of alkaloids (drugs): The latter appear to fit completely inside or at the mouth of the receptor core, whereas peptides appear to bind to extracellular loops. Differential binding sites allows for differential ligand effects, influencing conformation and subsequent responses. Opioid receptors differ from many other receptor systems in that a large number of endogenous ligands interact with a small number of receptors, potentially limiting the precise control that characterizes most systems.²⁷ However, mechanisms whereby distinct responses to endogenous opioids might occur despite the lack of apparent substrate specificity include differences in duration of action, activation of multiple receptors or their heterodimers, production of endogenous opioids with unique activation profiles (perhaps in part through differential responses at the level of intracellular signaling), and differences in intracellular trafficking of the receptors themselves (see the discussion of adaptation). These mechanisms offer means by which pharmacologic interventions might be designed to generate the desired opioid response.

Opiate receptors occur throughout the body but are in high density in the dorsal horn of the spinal cord, where they are responsible for modulating pain reception. The NC/OFQ system is distributed in the hippocampus, cortex, and selected sensory sites and is responsible for complex behavior.²⁷ Peripherally, receptors also are located in the spleen, kidney, intestine, vas deferens, and retina.³¹ These include effects on drug reward and reinforcement, stress responsiveness, and feeding behavior. The system is closely related to stress processes. Lester and Traynor³² demonstrated in vitro that the effect of commonly used opioids correlated to their binding of opioid or NOP receptors throughout the brain and spinal cord of dogs.

Central Nervous System

The major disadvantages of opioids reflect general CNS depression, including dose-related respiratory depression and, to a lesser degree, cardiac depression. CNS depression tends to preclude the use of opioids in syndromes such as shock, severe cranial trauma, and diseases associated with respiratory compromise. Synthetic opioids were designed to induce analgesia with minmal undesirable side effects (i.e., sedation, respiratory depression).

Sedation is common with opioid use, depending on the drug and its target receptors, and the species (i.e., MOP and to a lesser degree KOP receptor stimulation in the dog) and can be a disadvantage or an advantage, depending on the clinical situation. Species differences in response to the sedative effects of opioids can be profound. “Morphine mania,” typical of that described in cats is manifested as dysphoria and psychomotor activity. It may reflect sigma receptor stimulation for selected opioids or simply may reflect overdosing. Opioids tend to cause release (rather than inhibition) of some neurotransmitters (e.g., dopamine, acetylcholine), and this may occur in cats with high doses. When opioids are used in combination with phenothiazine derivatives (which block many of the neurotransmitter actions), the incidence of adverse effects appears to be reduced in cats. Dogs also are subject to dysphoria, as is exemplified by a series of three cases reported by Hofmeister and coworkers.³⁹ Dysphoria was manifested as vocalization that began within 5 minutes of administration of either hydrocodone, morphine, or fentanyl in association with postoperative analgesia. Clinical signs abated with administration of naloxone.

Respiratory depression usually is the cause of death in humans who succumb to opioids. The mechanism is reduction in the responsiveness of the brainstem respiratory centers to carbon dioxide, the primary stimulation for respiration. Centers that regulate respiratory rhythm are also depressed.²⁹ Depression, manifested as a slow respiratory rate, is discernible at doses lower than those associated with sedation and increases as the dose increases. Rate, tidal volume, and minute volume all decrease. Respiratory depression is, however, rarely a clinical concern except in cases of overdose or in the presence of pulmonary dysfunction in cases of standard doses.²⁹ Opioids should be used cautiously in patients with compromised respiratory function. Patients may appear to be handling the drugs well but in fact may be using compensatory mechanisms such as increased respiratory rate.²⁹ Concentrations of CO₂ may be increased, and respiratory centers may already be less sensitive to CO₂. The administration of an opioid may be dangerous in such situations. Use of opioids in the pregnant animal can lead to marked respiratory depression in the developing fetus, with little to no effect on the mother, because of the underdeveloped blood–brain barrier in the fetus.²⁹

Opioids have precipitated attacks of asthma in human anesthetized patients, probably as a result of histamine release. The importance of this in animals is not clear.

Opioids increase intracranial pressure as a result of increased concentrations of CO₂ and cerebral vasodilation. Cerebrospinal fluid (CSF) pressure also increases.²⁹ These effects may be exaggerated after head injury. The effects on intraocular pressure are not clear and may vary with the species. In humans accommodation is increased, with a decrease in intraocular pressure. Opioids cause miosis in humans, and mydriasis occurs in some species. Mydriasis may impede vision in cats; cats will be sensitive to light until mydriasis resolves.⁴³

Convulsions occur in some species when opioids are administered in high doses. Mechanisms probably include inhibition of GABA and may be more likely with morphinelike drugs.²⁹ The convulsant effects of some opioids can be reversed by naloxone,²⁹ suggesting that drugs with antagonistic actions (e.g., butorphanol, buprenorphine) at some receptors might be preferred in the patient having seizures. The impact of opioids on epileptic patients is not clear.

Morphine and related opioids directly depress the cough center at concentrations lower than that required for analgesia. Respiratory depression and cough suppression do not appear to be related. Thus antitussive opioids do not necessarily cause respiratory depression.²⁹

In contrast to depression, opioids directly stimulate the chemoreceptor triggering zone and thus may cause nausea and vomiting. Individual differences in the emetic response to opioids are marked in humans, but the amount of variability is not clear in animals.²⁹ In human patients in whom opioids cause emesis, after subsequent administration, opioids act as antiemetics, blocking further response by the chemoreceptor trigger zone to opioids. Actions at the vestibular apparatus may also be responsible for emesis. The emetic effects that typify administration of opioids as sole agents do not typically occur in the postoperative, sick, or pain-ridden patient. Butorphanol has been used in some species as an antiemetic to control vomiting induced by cisplatin.⁴⁴ In the awake dog, at doses associated with sedation, MOP agonists (morphine, fentanyl, and methadone) prevent emesis induced by apomorphine and copper sulfate. At lower doses morphine was able to cause emesis. KOP agonists also blocked the apomorphine emetic response at sedative doses. Antiemetic effects were blocked by naloxone. In contrast, DOP agonists did tend to cause emesis, leading the authors to conclude that delta receptor stimulation is associated with emesis, but MOP and KOP receptor stimulation with antiemesis.⁴⁵ Nausea, vomiting, and salivation are associated with morphine or hydromorphone administration in cats. Buprenorphine, butorphanol, or meperidine are unlikely to be associated with these side effects.^43,46 Decreased appetite may occur after several days of continuous opioid treatment in cats.⁴³ Urinary retention has been described for MOP-active opioids in animals, although effects may vary with species.⁴⁷ Mechanisms may reflect a centrally mediated vasopressin effect on the kidneys. Alternative mechanisms may include opioid-induced hypotension, which is probably more likely with morphine. Anderson and Day⁴⁷ demonstrated that both fentanyl and morphine decreased urine output in normal and traumatized dogs.

Both long- and short-term use of opioids has paradoxically been associated with hyperalgesia. A biphasic, dose-dependent effect has been described and may reflect an initial increase in excitatory neurotransmitters. The hyperalgesic effects of opioids may involve NMDA receptors. Opioids appear to increase NMDA neural currents, providing a rationale for the combination of opioids with NMDA antagonists for enhanced analgesia. Several types of opioid hyperalgesia have been described both experimentally and clinically. At low doses, probably a reflection of disinhibition, intrathecal opioids induce an excitatory effect, which is otherwise masked with the sedation typical of larger doses. In contrast, very high intrathecal doses stimulate allodynia. With single-dose opioids, increased nociception may appear as the opioid effect diminishes. This may be manifested as an increased postoperative need for analgesia in patients receiving short-acting opioids intraoperatively. Reversal of opioids with naloxone often produces marked hyperalgesia. Hyperalgesia has also been associated with chronic use; among the clinical sequelae might be marked hyperalgesia with opioid withdrawal.^48-50

Changes in body temperature reflect altered thermoregulatory response. Hypothermia is more common in dogs, whereas hyperthermia is more common in cats.⁵⁶ Opioids most commonly associated with hyperthermia in the cat include hydromorphone (at 3 times the dose) and meperidine; buprenorphine is unlikely to be associated with hyperthermia in cats.^43,57,58 Experimentally, hydromorphone (1 mg/kg intravenously) increased body temperature (range 104^° to 108^° F) 1 to 5 hours after anesthesia. Doses at or less than 0.05 mg/kg had no effect, whereas 0.1 mg/kg was associated with increased skin temperature.

Prolonged exposure of opioid receptors to ligands results in adaptation to their presence at multiple cellular levels.³⁰ Cellular tolerance is translated to animal adaptation manifested as tolerance, physical dependence, sensitization, and withdrawal. However, their emergence is not a reason to forgo opioid use. Tolerance refers to a decrease in drug efficacy associated with repeated administration. Acute tolerance (i.e., tachyphylaxis) may reflect short-term receptor desensitization, perhaps owing to phosphorylation of MOP or KOP receptors. Long-term administration of MOP ligands causes superactivation of adenylyl cyclase, the mechanism traditionally assumed to be associated with long-term tolerance. Different ligands appear to initiate acute or chronic tolerance through different mechanisms. Examples include internalization of receptors (MOP or DOP but not MOP), truncation of receptors or other changes. Adaptation by one response may mitigate adaptation by downstream responses; indeed, this sequela (i.e., failure to initiate downstream adaptations) has been proposed as a mechanism of failed desensitization that characterizes some drugs.²⁷ Accordingly, the ability of different ligands to initiate tolerance differs. Changes in nitric oxide or neurotransmitters or their pathways have been implicated as contributors to the development of tolerance.²⁹ Tolerance will most likely develop for analgesia, euphoria, sedation, respiratory depression, nausea or vomiting, and suppression of cough.

Physical dependence occurs when continued administration is necessary to prevent clinical signs characteristic of withdrawal. Physical dependence on opioids occurs as exogenous opioids replace endogenous opiates. Opioids affect numerous physiologic systems that become imbalanced before drug administration. A new balance or equilibrium is established in the presence of the drug,²⁷ and abrupt discontinuation of the drug requires rapid readjustment to a new equilibrium, predisposing the patient to withdrawal. Care must be taken to discriminate between dependence, which is associated with physical withdrawal, and addiction.

Addiction to opioids may reflect attenuation of the inhibition of dopamine release in the nucleus accumbens. This region of the brain has a central role in the reward circuit; dopamine promotes desire. The prevalence of addiction in humans using opoids is high. ⁵¹ Abrupt discontinuation of opioids after chronic dosing leads to symptoms of withdrawal, with the severity and duration reflecting the rate of onset and clearance of the opioid. Withdrawal is generally manifested as sign opposite to the original effects caused by the drug and reflects CNS hyperarousal as readaptation to the absence of the drug occurs.²⁷ Symptoms in humans include nausea and diarrhea, coughing, tearing, yawning, sneezing, rhinorrhea, profuse sweating, twitching muscles, abdominal and muscle pain and cramps, piloerection, and dysphoria. Hyperthermia, tachypnea, tachycardia and hypertension also may occur. Pharmacokinetic variables may be helpful in the prediction of withdrawal.²⁷ Morphine dependence has been described in dogs.⁵² Mixed agonists–antagonists or partial agonists are associated with the least risk of dependence; among the opioids studied, buprenorphine appears to be the least likely to cause dependence in dogs.⁵³ Opioids can be discontinued in drug-dependent human patients without causing signs typical of withdrawal by decreasing the dose 25% to 50% every couple of days. In human patients suffering from withdrawal, clonidine, an α₂-adrenergic agonist, minimizes the autonomic symptoms of opioid withdrawal.²⁷

It is the advent of tolerance and physical dependence that has led to the scheduling or classification of most opioids as potential substances of abuse. The actual class designated varies among the states. Pure agonists tend to be scheduled in class II or III; mixed or partial agonists tend to be scheduled in class IV or V, depending on the abuse potential. Both butorphanol and burpenorphine have been rescheduled from class V to class IV.

Cardiovascular System

The effects of opioids in the myocardium have been described. Cardiac depression (particularly bradycardia) is caused by selected opioids; pretreatment with atropine can reduce the incidence. The risk of hypotension is increased with drugs that also cause histamine release.

Those opioids most likely to be associated with histamine release are morphine (but not oxymorphone) and meperidine.^40,41 Histamine release caused by buprenorphine has been demonstrated only in vitro; fentanyl, oxymorphone, and butorphanol do not cause histamine release in dogs (as reviewed by Guedes and coworkers).⁴¹ Hydromorphone was associated with an apparent type I allergic reaction when used preoperatively in a dog (Chapter 2). However, species differences should be anticipated. Histamine antagonists (H₁) partially block morphine-induced hypotension; naloxone completely blocks it in human patients.^29,42 Fentanyl and its congeners are less likely to cause hypotension associated with surgery in part because they do not cause histamine release.²⁹ Volume replacement should be instituted before administration of morphine derivatives that causes histamine release in patients that are at risk for hypovolemia. Allergic phenomena in general (including bronchoconstriction) and those manifested in skin may be exacerbated with opioid use if morphinelike drugs are used. In humans opioids cause vasodilation of cutaneous vessels, probably because of histamine release. Pruritis may occur, in part because of histamine but also because of direct effects on neurons.²⁹

Other Effects

Ureteral tone may increase and the voiding reflex of the bladder may be diminished as a result of the increased tone of the external sphincter and increased volume of the urinary bladder. Morphine appears to have antidiuretic effects.²⁹ Inhibitory effects on uterine tone may prolong labor; hyperactivity induced by oxytocin can be normalized with morphine. Neonatal health may be impaired by use of morphinelike drugs during parturition; the neonate appears to be particularly susceptible to respiratory depression induced by opioids, in part because of a poorly developed blood–brain barrier.²⁹

Opioids appear to inhibit cytotoxic activity of natural killer cells. Selected opioid compounds, however, appear to enhance macrophage and killer cell activity, possibly through a novel opioid receptor (see Chapter 31).²⁹

Drug Interactions

The depressant effects of opioids can be exacerbated by other CNS depressants. Several CNS depressants prolong the effects of opioids, including the phenothiazines and tricyclic antidepressants (TCAs). Although some phenothiazines may reduce the amount of opioid necessary for analgesia, others may increase the amount of opioid.²⁹ Among the drug interactions associated with opioids are those at the receptor level (pharmacodynamic) – a interaction often intended for the purposes of reversal of undesireable clinical effect. These interactions are discussed with antagonists.

Disposition of Morphine

In contrast to cats, parenteral and epidural administration of morphine is generally well established in dogs and multiple studies are available to describe its disposition; several studies also address its metabolites (Tables 28-3 and 28-4; see also Table 28-1). Numerous studies using variable preparations, routes and doses in dogs are summarized in Table 28-4 and below. ^65-75 Multiples studies have focused on more convenient routes of administration (other than IV), including oral (including extended release products), rectal and epidural. Much of the IV data has been generated in studies that focus on alternate routes of delivery or in the context of determining analgesic concentrations.

Table 28-3 Drugs with Potential Epidural Use in Dogs or Cats²²⁷

Peak morphine concentration has been determined in several studies after intravenous, IM and SC administration. Morphine clearance appears to be dose dependent in the dog, with an elimination half-life of 83 ± 8 minutes at doses ranging from 7.2 to 7.7 mg/kg versus 37 ± 13 minutes at 0.019-0.07 mg/kg intravenously.⁶⁸ Bioavailability of intramuscular morphine is high (119%) but variable (57% to 161%), and elimination half-life approximates 90 minutes.

Studies focusing on the nociceptive concentration of morphine revealed an effective 50% concentration of 29.5 ± 5.4 ng/mL in dogs.⁷² The authors indicated that an intravenous infusion of 0.15 mg/kg/hr or multiple doses of 0.5 mg/kg every 2 hours for 3 doses provided significant nociception. The short half-life of morphine may require administration as a CRI in dogs. Because steady state may require 3 to 5 hours, a loading dose may be indicated. No differences in either analgesia or adverse reactions were detected in dogs undergoing exploratory laporatomy treated with morphine administered either intramuscularly (1 mg/kg at induction and extubation and every 4 hours thereafter) or CRI (0.12 mg/kg/h). Steady-state serum morphine concentration after CRI was 30 ± 2 ng/mL. The power to detect a significant difference was not addressed.

Morphine has been studied in dogs subcutaneously prepared as a solution and as a sustained-release gel (N,O-carboxymethylchitosan polymer) product⁶⁹ given either subcutaneously or orally⁷⁰ (it is not clear if the sustained oral preparation is the same as that approved for use in humans). After subcutaneous injection at 1.2 mg/kg, peak concentrations for the solution and sustained product were 488 and 180 ng/mL, respectively, with time to peak concentrations being 10 and 55 minutes, respectively. Variability among animals was large. Elimination half-life was 79 (solution) and 108 (gel) minutes. The duration of analgesia was reported to be between 1 and 2 hours but appeared longer clinically. Emesis should be anticipated.⁵⁶

Oral administration of morphine in humans generates effective systemic concentrations, although first-pass metabolism requires oral doses that are two to six times parenteral doses. A sustained-release oral morphine preparation (Morphelan) has been approved for use in humans in the United States. The product is designed for both immediate and extended release. Because several days are required for steady-state kinetics to be reached, an initial loading dose of twice the intended subsequent daily dose has been recommended to achieve steady-state concentrations within 2 days of initiation of therapy.⁷¹ However, the role of oral administration of morphine in providing effective analgesia in dogs is not clear. The formation of M6G after oral administration of approximately 1.6 mg/kg, morphine sulfate extended release in dogs was previously discussed. In their study, KuKanich and coworkers ⁶⁶ compared the disposition of intravenous morphine (morphine sulfate 0.5 mg/kg) and orally administered morphine as a commercially available extended-release tablet (1.6 ± 0.1 mg/kg) in Beagles (n=6) (see Table 28-4). One dog vomited after oral administration. Oral absorption of the extended-release tablet was poor (5%) and erratic, achieving a C_max of morphine of 5.0 ng/ml (variability not provided); this low concentration presumably reflects poor oral absorption rather than high first pass metabolism in that M6G was not detected. This compares to an extrapolated (after distribution) peak concentration of 60 ± 10 ng/ml after IV administration of 0.5 mg/kg. Accordingly, for effective analgesia, a dose of 0.5 mg/kg intravenously every 2 hours was recommended.

More recently, Aragon and coworkers⁷⁴ reported the disposition of morphine (1 and 2 mg/kg) as either an immediate or oral sustained-release product in dogs (n=14). Although drug concentrations were detectable for 24 hours, concentrations were considered too low and too variable to be of clinical use.

Rectal administration may be a viable alternative route of of administration in dogs. After rectal administration in dogs, morphine is approximately 20% bioavailable,⁷⁵ achieving peak concentrations of 28 ng/mL when 2 mg/kg is administered as a solution and 20, 21, and 51 ng/mL when administered as a suppository at 1, 2, and 5 mg/kg, respectively. Elimination half-life (minutes) ranged from 65 (solution) to 98 (suppository) (see Table 28-4). Vomiting was common (five of six animals) with a 5-mg/kg rectal suppository.

Morphine was studied in dogs after intramuscular and rectal administration, the latter as either a solution or suppository (see Table 28-4). Median peak concentrations were reported at 5 minutes. The duration that drug concentrations were above a target of 20 ng/mL were 120 minutes for the intramuscular dose, 5 to 30 minutes for the rectal solution, less than 40 minutes for the low (1 mg/kg) and medium (2 mg/kg) suppository dose, and 120 minutes for the high suppository dose (5 mg/kg morphine sulfate).⁷⁵

Morphine also has been studied in dogs after epidural administration of the solution (0.1 mg/kg)⁷⁶ and as a sustained-release encapsulated preparation (experimental; 10 and 30 mg in 3 mL saline).⁷⁷ Both appear to prolong analgesia while avoiding side effects. King and coworkers⁷⁸ demonstrated that epidural morphine (0.07 mg/kg, containing 0.1% sodium metabisulfate) was not associated with histologic damage to the spinal cord in dogs (n=16). Morphine has been studied in Beagles after repetitive epidural administration (30 mg at 10 mg/mL) as a sustained-release multivesicular liposome preparation (DepoFoam drug delivery system).⁷⁹ Adverse reactions were limited to moderate, transient behavioral and physiologic changes, consistent with morphine administration, at each injection. A modest effect on cord histopathology was not associated with changes in CSF or neurologic signs.⁷⁹

Morphine also has been studied intravaginally in humans. When this route was used, pain was effectively reduced in one patient, although close monitoring was necessary.⁸⁰ Morphine has been studied in both dogs and cats after transdermal administration as a pluronic lethicin organo (PLO) gel. Drug was not quantifiable in either species when administered at 1.6 to 2 mg/kg.⁶⁵

Adverse Events and Drug Interactions

Adversities of morphine were discussed as a class. Morphine appears to cause more sedation than transdermal fentanyl when used as a postoperative analgesic.⁷⁶ Guedes and coworkers⁴¹ demonstrated that morphine was associated with variable histamine release in conscious dogs at 0.3 and 0.6 mg/kg load followed by 0.17 or 0.34 mg/kg/hr, although the difference from placebo was statistically significant only for the higher dose. Flushing of the skin occurred within 5 minutes of 60% and 100% of dogs receiving the lower and higher dose, respectively. Although no adverse cardiovascular events were detected, adversities might be anticipated in at-risk patients.

Intrathecal administration of morphine at 0.15 mg/kg has been associated with a number of adverse effects in dogs, including bradycardia, hypotension, myoclonus, and urinary retention (as reviewed by Novello and coworkers).⁸¹ However, Novello and coworkers⁸¹ demonstrated in clinical cases undergoing cervical or thoracolumbar laminectomy that a low intrathecal dose of morphine (0.023 to 0.034 mg/kg) administered at the lumber level 25 to 65 minutes before surgery reduced the dose of fentanyl (1.2 μg/kg/hr) needed to maintain a steady-state heart rate and arterial blood pressure when compared with a placebo group (4.2 μg/kg/hr fentanyl infusion). Ketamine was also administered to all animals (0.5 mg/kg every 60 minutes starting 10 minutes before surgery). Induction and fentanyl protocols differed among animals in the study; it is not clear if both methods were distributed similarly among treatment groups.

Vomiting is relatively consistently associated with morphine administration in dogs. Vomiting occurred after intravenous (0.5 mg/kg) or intramuscular (1 mg/kg) administration in one study (five of six dogs) at a mean morphine concentration of 66 ng/mL, and heart rates were significantly lower.⁷⁵Drug interactions with morphine appear to be limited. KuKanich and Borum⁷³ found morphine disposition in Greyhounds was not affected by ketoconazole. The simultaneous combination of morphine with the pure antagonists naltrexone did not affect the disposition of either drug in dogs.⁸²

Therapeutic Use

Morphine has been used in the dog and cat effectively as a premedicant or a perioperative analgesic⁵⁶ using a variety of routes. Lucas and coworkers⁸³ found no differences in the analgesic efficacy of morphine when administered as a CRI (0.12 mg/kg/hr) or intramuscularly at intubation, extubation, and every 4 hours in dogs (n=20) undergoing exploratory laparotomy. Morphine can be administered safely epidurally, although preparations made specifically for epidural administration are preferred. Epidural administration appears to be effective when given either epidurally^84,85 or locally (intraarticularly)⁸⁴ for control of some orthopedic pains.

The postoperative analgesic effects of epidural morphine (0.1 mg/kg) or morphine combined with medetomidine (0.005 mg/kg) have been compared in dogs (n=6/group) undergoing cranial cruciate repair. No differences were detected in control of pain (power of the study not reported), although the combination drug consistently had a lower numeric score. Increasing sample size may have resulted in statistical signification. Likewise, no differences were detected in plasma morphine concentrations between the two treatment groups (peak concentrations of approximately 4 ng/mL).⁸⁶ The effect of morphine on healing corneal ulcers has been studied in dogs after administration of an ophthalmic preparation. Both MOP and DOP receptors have been identified in the canine cornea. Topical 1% (but not 0.5%) morphine provided good corneal analgesia with no apparent adverse effects.⁸⁷

Morphine (0.2 mg/kg subcutaneously) was compared with buprenorphine (0.02 mg/kg subcutaneously) and methadone (0.2 mg/kg subcutaneously) in cats (n=8) subjected to mechanical and thermal stimulation. Morphine provided the best analgesia.¹²⁶

Pharmacokinetics and Pharmacodynamics

Fentanyl has been studied in dogs and cats using several different methods of delivery. Using a thermal threshold response, fentanyl concentrations greater than 1 ng/mL have been recommended for therapeutic analgesic effects in cats.¹⁰¹ A dose of 10 μg/kg intravenously provided analgesia (measured by thermal response) for approximately 2 hours in cats.

In normal dogs the elimination half-life approximates 3 hours and the volume of distribution 10 L/kg (see Table 28-4); μg/kg was 1.5 and 2.8 ng/mL, respectively; elimination half-life was approximately 2.5 hours.⁶⁵ In normal cats the elimination half-life also approximates 3 hours, but volume of distribution appears to be smaller.^102. A half-life approximating 2.5 to 3 hours in either species potentially supports the use of a CRI, although a loading dose is indicated. An infusion rate of 0.4 μg/kg/min after a loading dose of 5 μg/kg is recommended in cats. For dogs, fentanyl has been studied after an intravenous loading dose, and an intravenous loading dose (10 μg/kg) followed by CRI (10 μg/kg/hr) for 1 to 3 hours (see Table 28-4).¹⁰³ Concentrations declined to below 1 ng/mL by 20 minutes after the intravenous loading dose and reached 1 ng/mL again by about 2 hours, with steady-state concentrations at 1, 3, and 4 hours of infusion being 1.0 ± 0.7, 1.11 ± 0.4, and 1.25 ± 0.7 ng/mL, respectively. Variability in concentrations was marked. The elimination half-life of fentanyl after the CRI was discontinued was 151 ± 61 (2.51+1 hours) to 182 ± 68 (3+1 hours).

Fentanyl is among the drugs for which alternative routes of administration have been developed,¹⁰⁴ several of which have been studied in dogs and cats. Transdermal delivery of 10 or 30 μg/kg as a PLO gel yielded no detectable plasma fentanyl nor any analgesia in cats in one study.¹⁰¹ However, a second study (unpublished data by author) found a C_max of 8.8 ± 5 ng/mL at 10 ± 19 hours after transdermal delivery of a PLO gel (5 mg/mL) to cats (n=3) at a mean dose of 0.08 ± 0.03 mg/kg (80 ± 30 μg/kg). Potentially therapeutic concentrations persisted throughout the duration of the study: concentrations of 4 ± 5 ng/mL were measured at 61 hours. In contrast to cats, fentanyl was not detected in dogs after administration of 0.88 mg/kg (880 µg/kg) as a transdermal PLO gel.⁶⁵

Fentanyl (2 μg/kg) has been studied after oral and intranasal administration in cats. C_max (ng/mL) 5 minutes after oral administration was 0.96 and 2 minutes after intranasal administration, 1.48 ng/mL.¹⁰¹ Although not yet studied in dogs or cats, nn humans, after administration as an aerosol (50 μg/mL solution aerosolized for 5 minutes), fentanyl concentrations were approximately 25% of that achieved after 100 μg intravenously (1.4 ng/mL versus 4.4 ng/mL, respectively), suggesting aerosolization as a possible route of administration.¹⁰⁵

Fentanyl has been available for about 10 years as a transdermal drug delivery system intended for control of pain (Figure 28-4). Issues surrounding transdermal drug delivery in animals have been well reviewed. ¹⁰⁶ The transdermal patch system has proved safe and effective for delivery of fentanyl to both dogs and cats. The transdermal drug delivery system provides slow, continuous drug delivery that is intended to provide a fairly constant plasma fentanyl concentration. Peak and trough concentrations, which might cause toxicity and therapeutic failure, respectively, are thus avoided. The fentanyl transdermal patch system has been approved for control of cancer pain in humans. The system consists of a patch with an adhesive layer that attaches to the skin. A “release” membrane controls the rate of release of drug from the reservoir. Because the amount of drug that is released is proportional to the size (area) of the patch,¹⁰⁷ the dose delivered is the amount of drug (in μg) released per unit time (hour). The system is available in four sizes: 25, 50, 75, and 100 μg/hr. Drug delivery through the stratum corneum can be influenced by a number of factors that vary among patients. These include thickness of the stratum corneum and epidermal layer, skin appendages (e.g., number of sweat glands), body temperature, hydration status, and diseases that might alter the permeability of the stratum corneum.

Figure 28-4 The transdermal fentanyl patch system. The dermis is the site of microcirculation and thus drug absorption.

Species difference mandate that the fentanyl patch system be studied in the target species before use. Peak concentrations occur within 24 hours in the dog and can occur in as few as 3 to 5 hours in the cat. The commercially available fentanyl patch has been studied in both dogs^102,108,109 and cats.¹¹⁰ With the 50 μg/hr patch, drug delivery approximated 37 μg/hr (range of 13.7 to 59.8 μg/hr) in dogs, and the patch was found to be an effective means of constant, slow drug delivery.¹⁰⁸ A comparison of drug delivery from different patch sizes found the average fentanyl concentration from 24 to 72 hours to be 0.7 ng/mL (50 μg/hr), 1.4 ng/mL (75 μg/hr), and 1.2 ng/mL (100 μg/hr) in dogs weighing approximately 20 kg.¹⁰⁹ The elimination (disappearance) half-life ranged from 3.6 hr (50 μg/hr) to a low of 2.5 hr (100 μg/hr). Variability among dogs was marked, making duration of analgesia difficult to predict in individual animals.

Fentanyl delivery has been studied after administration of a 25 μg/hr patch in cats.¹⁰² Delivery was approximately 8.5 μg/hr (36% of the predicted) with mean steady-state concentrations of 1.58 ng/mL being achieved at 12 to 100 hr; the maximum concentration of 3.6± 0.8 was achieved at 44±12 hr. Concentrations dropped below 1 ng/mL in four of six cats within 12 hours of patch application and in all cats within 36 hours. The amount of drug delivered through a fentanyl transdermal patch can be reduced by decreasing the surface area contact with the skin. The most accurate approach probably is removal of only a portion of the protective lining cover of the patch, although the proportion of drug delivery will not be exactly proportional to the amount of lining removal. Folding half of a 25 μg/hr patch in cats yielded mean peak concentrations of 1.14 ± 0.9 ng/mL compared with 1.8 ± 0.9 ng/mL (38% reduction) for full patch exposure. ¹¹⁰

Robertson and coworkers¹⁰¹ determined the pharmacodynamic response of fentanyl either intravenously or as a PLO (10 μg/kg) applied to the inner pinna. Plasma fentanyl concentrations greater than 1 ng/mL were required for effective analgesia. The only value reported in the study was C_max (rather than extrapolated γ intercept) for fentanyl after intravenous administration. However, mean concentrations appeared to be above 1 ng/mL for approximately 60 minutes after intravenous administration. Elimination half-life estimated from the plasma drug concentration versus time curve approximated 15 to 30 minutes. Cats tolerated fentanyl administration well, with most cats exhibiting euphoria. Fentanyl was not detected after PLO administration when administered at 30 μg/kg.

In this same study, Robertson and coworkers¹⁰¹ studied fentanyl after oral or intranasal administration of 2 μg/kg, yielding C_max values of 0.96 and 1.48 ng/mL at 5 and 2 minutes, respectively. Cats receiving the drug orally or intranasally salivated profusely and were difficult to restrain. Thermal antinociception differed from saline control for approximately 2 hours.

The rate of drug release from the patch and skin can vary with environmental factors that alter delivery through the stratum corneum. Most notable is temperature; fever or a heating pad will increase drug movement. Body temperature was shown to influence fentanyl concentrations when delivered using a fentanyl as a transdermal patch in cats undergoing isoflurane anesthesia. Peak concentrations were 1.8 ± 0.6 ng/mL in normothermic patients compared with 0.6 ± 0.3 ng/mL in hypothermic cats.¹¹⁴ In another study, experimentally induced hypothermia (35^°C) was associated with a decline in plasma fentanyl concentrations of 33% (0.6 ng/mL) compared with that at baseline (1.8 ng/mL).¹¹⁷ Patches had been applied 24 hours before anesthesia. In contrast, hyperthermia may increase the risk of fentanyl overdose.¹¹⁵

Therapeutic Use

Only a few clinical trials have compared non-transdermal fentanyl to alternative opioids. Anderson and Day⁴⁷ compared the effects of morphine (0.12 mg/kg/hr), fentanyl (3 μg/kg/hr) administered by CRI with those of placebo on urine output in healthy (n = 23) and traumatized (n = 18) patients. All animals received a similar rate of fluid infusion. Both fentanyl and morphine decreased urine output in both the trauma and healthy groups (approximately 0.7 to 0.76 mL/kg/hr in all groups compared with 1.2 mL/kg/hr in the untreated group).

Several studies have demonstrated efficacy of the fentanyl transdermal patch system in cats.^110,111,113 Fentanyl administered as a 25 μg/hr patch was effective on controlling pain associated with ovariohysterectomy in cats.¹¹¹ Fentanyl was more effective than butorphanol (0.4 mg/kg intravenously) in cats undergoing ovariohysterectomy or onychectomy.¹¹³ A number of studies have documented analgesic efficacy of transdermal fentanyl in dogs. Analgesia did not differ from produced by intramuscular oxymorphone (0.05 mg/kg), but fentanyl was associated with less sedation when used in canine patients undergoing ovariohysterectomy.¹¹² Robinson and coworkers⁷⁶ compared the efficacy of transdermal fentanyl (100 μg/hr; placed 24 hours before surgery) with that of epidural morphine (0.1 mg/kg) at induction of anesthesia in dogs (n=10) undergoing major orthopedic surgery. Fentanyl provided superior analgesia. The fentanyl patch system was equal to or better than epidural morphine (0.1 mg/kg) in canine patients undergoing orthopedic surgery⁷⁶.

The fentanyl transdermal patch appears to be a reasonable alternative to pain management for dogs and cats, but with several caveats. Patches can cause irritation. Because the time to peak therapeutic concentrations of the drug may be up to 24 hours,¹⁰⁹ the need for shorter-term control of pain (e.g., for postoperative pain) should be anticipated. Patches should be applied 12 to 24 hours before the anticipated need for analgesia. Likewise, after the patch is removed, a similar 12- to 24-hour period must elapse before the drug is no longer detectable in the patient. Analgesia is provided for 24 to 72 hours, although variability among animals can be marked, making use of the patch for the individual animal less predictable. Intravenous fentanyl (30 μg/kg has been recommended for dogs) can be given at the time of patch application as a loading dose for patients in need of immediate analgesia. Other opioids also can be given intravenously, intramuscularly, or subcutaneously until the patch is effective. Although a pure opioid (morphine, oxymorphone) is probably preferable, a mixed or partial drug (butorphanol or buprenorphine, respectively) can be given. Because both of the latter drugs have some MOP antagonist effects, however, the MOP effects of fentanyl from the patch will be blocked until the antagonist is eliminated. This may be more problematic with buprenorphine, which has a very tight affinity for MOP receptors.^113a

The transdermal delivery of fentanyl appears to be very useful in small animals experiencing pain, including postoperative pain, cancer pain, and pain associated with trauma. The patches can be dosed as follows on the basis of body weight: 5 to 10 kg: 25 μg/hr; 10 to 20 kg: 50 μg/hr; 20 to 30 kg: 75 μg/hr; and more than 30 kg: 100 μg/hr.¹¹⁵ The dose of fentanyl administered by a transdermal patch can be decreased in animals ≤5 kg by folding the adhesive membrane such that only a portion of the patch is exposed to the skin or by cutting away half of the seal (but not the patch itself). However, variability in concentrations will complicate effective use. Exposure of the second half of the patch may not provide equivalent analgesia if applied to a different cat or after all fentanyl has been eliminated. For animals larger than 30 kg, multiple patches can be applied. Patches can be used on an outpatient basis. Heating pads should not be placed at the site of a patch. Patches generally are placed on the dorsum of the neck, which first must be clipped and dried. The patch must come in close, snug contact with the skin. A bandage should be applied to keep the patch in place and prevent the animal or people (especially children) from disturbing the patch.

If the patch is prescribed, veterinarians should apply the patch to, and on completion of analgesia remove it from, the animal rather than allow the client to do so. Clients should be warned that the patches are not approved for use in animals. Patches that are sent home with the patient should be collected when removed from the animal to minimize the risk of drug abuse by pet owners.¹¹⁶ Animals occasionally remove the patch and swallow it; however, the risk of drug toxicity is minimized because it is unlikely that the close contact necessary for drug delivery will occur between the mucosa and the patch. In addition, first-pass metabolism of any fentanyl that is absorbed will limit the amount of drug that reaches systemic circulation. This will not be true if the patch is chewed such that absorption through the buccal mucosa occurs.