20

Changing Concepts of Tolerance to Opioids: What the Cancer Patient Has Taught Us

Kathleen M. Foley Departments of Neurology, Neuroscience, and Clinical Pharmacology, Cornell University Medical College and Pain Service, Department of Neurology, Memorial Sloan-Kettering Cancer Center, New York, New York 10021

In the management of patients with pain and cancer, one of the controversial questions has been: To what extent does tolerance limit patients' ability to obtain adequate analgesia from opioid therapy during the course of their illness? This is a critical question because it is physicians' concern about the development of analgesic tolerance that has traditionally limited the chronic use of opioids in patients with chronic pain. The cancer patient with pain has served as a natural experiment to begin to study this question, and currently available clinical data correlated with new discoveries in the molecular mechanisms of tolerance have facilitated this discussion (8,9,11).

DEFINITIONS OF TERMS: TOLERANCE, PHYSICAL DEPENDENCE, AND PSYCHOLOGICAL DEPENDENCE

Tolerance is the term used to define the phenomenon in which an organism is less susceptible to the effect of a drug as a consequence of its prior administration. Acute tolerance is the term used to describe tolerance that develops very rapidly following either a single dose or a few doses given over a short period of time. Chronic tolerance describes the observation that drug administration over a longer period of time produces reduced drug effects. Cross-tolerance describes the observation that tolerance to one drug confers tolerance to another. Tolerance has also been described in such terms as associative or behavioral tolerance and nonassociative or pharmacological tolerance. Associative tolerance is best expressed with low doses of drugs at long interdose intervals and is readily modified by behavioral or environmental interventions. Nonassociative tolerance is best expressed with high doses of drugs at short interdose intervals and is not modified by behavioral or environmental interventions. It is thought that associative tolerance results from the learning of drug-environment associations, whereas nonassociative tolerance and dependence can be viewed simply as adaptive changes resulting from direct drug actions (6,7). These two types of tolerance can be dissociated in the laboratory. In the clinical setting, both behavioral and nonbehavioral factors are important in the development of tolerance. Physical dependence is the term used to describe the phenomenon of withdrawal when an opioid is abruptly discontinued or an opioid antagonist is administered. Both tolerance and physical dependence are predictable pharmacological effects seen in response to repeated administration of an opioid in both animals and humans. These effects are distinct from the behavioral pattern seen in some individuals and described by the term psychological dependence or addiction. Psychological dependence describes a pattern of drug use characterized by a continued craving for an opioid, manifested as compulsive, drug-seeking behavior and overwhelming involvement in drug procurement and use. There is now both animal and human data to suggest that the underlying mechanisms of tolerance, physical dependence, and psychological dependence are different and that these definitions have relevance to our discussions of the chronic use of opioids in patients with cancer and nonmalignant pain.

STUDIES OF TOLERANCE AND PHYSICAL DEPENDENCE IN ANIMALS

It is beyond the scope of this section to review the extensive literature on tolerance in animals. Numerous investigators have demonstrated that multiple factors can alter the development of tolerance in animals, demonstrating its behavioral and pharmacological aspects (1,4,23,27,44,50). These include the test paradigm, the learned behavior, and the duration of exposure to the opioid drug as well as pharmacokinetic (PK) and genetic factors. The classical paradigm of tolerance is described in Figure 1, in which rats are given continuous systemic (IV), intracerebroventricular (ICVT), or intrathecal (IT) administration of approximately equiactive doses of morphine (59). There is a decline in the level of analgesia observed over an ensuing 7-day period of infusion. The magnitude of tolerance induced by chronic exposure to opioids can be related to the amount of drug administered. It is this paradigm that has been used to suggest that a similar phenomenon would occur in humans. In this paradigm the pain stimulus is kept constant during the test period. Although these studies were typically done with morphine, it has now become clear that tolerance development is receptor selective (27,34,49,50,58,59). In the case of mu and kappa receptors, cross-tolerance does not develop. The degree to which cross-tolerance occurs with delta receptor agonists has not been fully assessed. Moreover, it has become apparent that various types of pain are more likely to activate certain opioid receptor mechanisms and therefore, the degree to which tolerance develops may also, in part, be related to the type of pain stimulus employed in the experimental paradigm. These data are summarized in Table 1 (56). 
Chapter 20 table 1: Contribution of mu, delta, and kappa opioid receptors to antinociception and cross-tolerance between receptors: a summary
 

+, general agreement of ``yes'' response; -, general agreement of ``no'' response; ± , equivocal results; may be dependent on species, level of neuroaxis, etc.; ?, has not been evaluated.
A. Endogenous activation
  Level of Neuroaxis   Cross Tolerance
Test Receptor involved Spinal Supraspinal Tolerance development Mu Delta Kappa
Electrical stimulation Mu + ? + + ? -
Foot Shock
  Intermittent Mu ? ? + + ? -
  Continuous Kappa ? ? + + ? ?
Swim stress Mu and delta ? ? + + ? ?
Arthritic Rat Kappa (Mu) + ? + - ? +
B. Exogenous activation
  Level of Neuroaxis   Cross Tolerance
Ligand Presumed Receptor Antinociceptor observerd Spinal Supraspinal Tolerance development Mu Delta Kappa
Morphine Mu + + + + + ±  -
DPDPE Delta + ±  ±  + ±  ? ?
U50,488H Kappa + ±  ±  + - ? ?
Dynorphin Kappa ±  ±  - + - ? ?
 
 

In an attempt to evaluate the development of morphine tolerance in a chronic pain model, morphine self-administration in rats with adjuvant-induced arthritis was studied (31). These experiments demonstrated that arthritic rats self-injected less morphine than pain-free rats ( Figure 2 ). Moreover, their dose remained stable over 29 days, in contrast to the pain-free rats who escalated their morphine doses. The investigators suggest that pain altered tolerance development in these animals. This model is very similar to the clinical observations in which stable opioid use has been demonstrated in patients with both chronic cancer pain and nonmalignant pain. This study, and several others, have suggested that the presence of pain has a significant influence on the development of tolerance and the pattern of its development.

STUDIES OF TOLERANCE AND PHYSICAL DEPENDENCE IN HUMANS

Studies to assess tolerance in humans were initiated by Light and Torrance in the late 1920s (30). These investigators studied addicts stabilized on doses of morphine and reported that their heart rate, blood pressure, and respiration were within normal limits. They were also the first to describe, in detail, the clinical abstinence syndrome. The major focus of research on the phenomenon of tolerance in the early twentieth century was on the relationship of tolerance to physical and psychological dependence. The commonly held pharmacological interpretation of opioid addiction was that the addict, in the setting of continuous use of opioids, would develop tolerance to the euphorigenic effects of these drugs (7,14,18,30,57). This would then force the addict to require larger doses of opioids to obtain a desired euphorigenic effect. This increase in the amount of opioid consumed was then further associated with the development of physical dependence, which became manifest when the addict discontinued the opioid. The discomfort associated with this syndrome of abstinence became a reinforcing property from which the addict learned to relieve the discomfort by reinstituting opioid use. In this setting, then, a true dependence was established with reinforcing drug-seeking behavior. Historically, then, tolerance and physical dependence were linked as manifestations of a single phenomenon. However, it has now been demonstrated that tolerance and physical dependence designate a constellation of phenomena involving a variety of mechanisms. Multiple attempts have been made to quantitate tolerance and to develop a ``tolerance index.'' Studies by Martin facilitated the development of ratios of doses of opioid required to produce an effect in nontolerant and tolerant subjects (34,35). In these studies, Martin obtained dose-effect relationships for several parameters; he observed the subjective effects of morphine and heroin prior to chronic administration of these drugs and, subsequently, the effective doses of morphine and heroin when the subjects were receiving the drugs chronically.

Because tolerance was widely considered to reinforce the street addicts' abuse of opioids, it was commonly thought that this phenomenon would occur in a comparable manner if opioids were to be used on a chronic basis in the management of patients with pain or other medical illness. Physicians warned patients to limit their use and even suggested that the drugs might not work for the patients when they ``really needed them.'' This strong negative opinion had an enormous impact on the clinical use of opioids. However, as we critically look at the phenomenon of tolerance in the clinical setting, it becomes apparent that studies assessing tolerance in this population of patients have been thwarted by numerous methodological difficulties. First of all, it is apparent that tolerance develops at different rates for each of the known opioid effects. Moreover, the rate of tolerance development for each of these effects has not been measured with any precision in patients receiving these drugs for analgesia.

The acute effects of opioids in humans are well known. These include analgesia, sedation, euphoria or dysphoria, respiratory depression, pupillary constriction, urinary retention, constipation, and a variety of alterations in endocrine function. There is sufficient data to support the concept that, during chronic use in humans, tolerance develops to most but not all of these acute opioid effects (9,11,19,52). In [chhumans, it appears that tolerances to mood-altering effects, respiratory effects, se[chdation, analgesia, emesis, and miosis develop more rapidly than tolerance to constipating effects. What is the most problematic, however, in carefully defining the development of tolerance to analgesia in patients receiving these drugs for pain relief, is that there is little control over the pain stimulus. Therefore discerning the relative contributions of a changing pain state from that due directly to tolerance has confounded the issue. For purposes of discussion, some of the important studies of clinical tolerance in humans will be briefly reviewed.

Studies at the Addiction Research Center (ARC) by Martin and Jasinski addressed the development of tolerance in groups of subjects (previous addicts) who had not received any opioid for 1 to 7 months prior to being admitted to the ARC (34,35). Controlled observations were obtained over a 2-month period before the subjects were made dependent on 240 mg/day of morphine over a 5-week period. These individuals were stabilized on morphine for 29 weeks, gradually withdrawn over a 3-week period, and then observed for a 30-week postwithdrawal period. While stabilized on morphine, subjects demonstrated a variety of changes in the autonomic nervous system, including an elevation in blood pressure, pulse rate, and rectal temperature and a decrease in respiratory rate. Comparable studies on such patient populations have been performed with methadone and with some of the mixed agonist/antagonist drugs. Following the period of chronic administration and then precipitous withdrawal, the signs and symptoms of abstinence were observed leading to the development of what is referred to as ``abstinence scores'' and a quantitative description of physical dependence. These data are summarized in Figure 3 and \ (34). These experiments defined the chronic effects of opioids on blood pressure, temperature, body weight, pupillary diameter, and respiratory rate. None of these addicts was assessed for pain or challenged with painful stimuli to assess the degree to which tolerance to analgesia occurred. What these classical studies showed is that tolerance does develop to opioid effects, but at varying rates. Studies in addicts by Kreek (28) showed that addicts maintained on methadone at a constant single daily dose did not develop tolerance to the opioid withdrawal prevention effects. These studies also showed that tolerance to the neuroendocrine effects of opioids occurred slowly, if at all, in patients maintained on methadone.

Taking a different tack, Houde et al. began to study the issue of tolerance to analgesic effects in a clinical pain paradigm=mthe cancer pain patient (21). Figure 5 demonstrates the effects of graded doses of morphine on pain relief in 10 patients studied 2 weeks apart during which time they received morphine chronically. A clear shift in the dose-response curve to the right demonstrates some degree of tolerance to the analgesic effects of morphine.

In a second study of tolerance and cross-tolerance, two groups of opioid-tolerant cancer patients with chronic pain were used (20). In each group, an initial double-blind relative potency comparison was made using 8 and 11.3 mg of metopon, 16 and 22 mg of morphine, and a placebo. This was followed by a period of approximately 1 week in which patients in one group were administered morphine while patients in the other group were given metopon on demand for pain. This period was followed by a second relative potency assay utilizing the same drugs and doses as the first. These results are summarized in Table 2 and Figure 6 and Figure 7 . Only seven patients completed the study of morphine tolerance and cross-tolerance and only six patients completed the study of tolerance and cross-tolerance to metopon. These results, therefore, cannot be looked on as definitive studies but, rather, as indicative of trends. They demonstrated that direct tolerance developed to both drugs and that cross-tolerance, while present, developed at a considerably slower rate. These studies help to point up the fact that cross-tolerance, although it occurred, was incomplete.

Chapter 20 Table 2:Tolerance and cross-tolerance to morphine and metopon based on relative analgesic potency assays before and after chronic administration of each drug in two groups of patients
Pre Chronic drug administration Post
 Equivalent analgesic dosesa  Morphine:  Equivalent analgesic dosesa
  MS: 10 mg  Mean daily dose = 77.3 mg  MS: 16.8 mg
  Me: 5.1 mg  Mean no. of days = 8.1 daysb  Me: 6.4 mg
   Metopon:  
  Me: 5.2 mg  Mean daily dose = 29.7 mg  Me: 17 mg
  MS: 10 mg  Mean no. of days = 6.8 daysc  MS: 22.8 mg
aEquated to effect of 10 mg in the prechronic administrative period. 
bAverage number of injections per day = 5.2. 
cAverage number of injections per day = 5.9. 
MS, morphine; Me, metopon. 
Further studies addressing the issue of cross-tolerance in humans to agonist and mixed agonist/antagonist drugs were demonstrated by Houde et al. (20) in studies in which a mixed agonist/antagonist drug, pentazocine, was administered to patients who were tolerant to morphine. Figure 8 shows both the design and results of these studies. In nontolerant patients, combinations of up to 80 mg of pentazocine with 8 mg of morphine produced additive analgesia. By contrast, in tolerant patients, the investigators observed antagonism rather than an increase in analgesia with the combination of 10, 20, and 40 mg of pentazocine with 8 mg of morphine. Pentazocine, 40 mg alone, produced no analgesia as compared to saline, and in some patients produced signs of an opioid abstinence syndrome. These investigators concluded that the antagonist/agonist potency ratio of the mixed agonist/antagonist drugs is altered in patients who are tolerant to and physically dependent on opioids as compared to patients who are not tolerant to opioids, pointing out the fact that using mixtures of mixed agonist/antagonist drugs in patients tolerant to mu agonist drugs diminishes their analgesic effects (20,25).

To summarize these clinical studies of analgesic tolerance, Houde et al. clearly demonstrated that tolerance develops to analgesia and that cross-tolerance is incomplete. These observations have been substantiated in several studies, including a report from Bruera et al. (2), who evaluated the cognitive effects of chronic administration of opioids to patients with advanced cancer who received increases in their opioid doses. Two populations were studied. One group (stable-doses group [SD]) had had no change in their opioid dose or type for up to 7 days; the other group (increased-doses group [ID]) had had a 30% increase in 3 days or less before the onset of the study. Both groups showed a decrease in pain and increased somnolence with more sedation and nausea in the ID group as compared to the SD group. This study again demonstrates the differential development of tolerance to analgesia and sedation.

In more recent studies, Inturrisi et al., using a compartmental model approach to characterize the PK and pharmacodynamics (PD) of methadone, have attempted to determine whether acute tolerance occurs to analgesia and sedation (22). Under study conditions, changes in plasma methadone and morphine concentrations can be directly correlated to changes in pain, pain relief, or sedation measured by the use of visual analogue scales or categorical scales. From these studies, it appears that PK factors are the predominant determinant of the intensity of opioid effects. Using PK/PD modeling, PD estimates can be made, including a Css50 that reflects the intrinsic sensitivity of the patient to the drug effect and gamma, the slope function. In Figure 9 , the concentration-effect plot during the infusion and washout in a patient with chronic pain receiving methadone is demonstrated. If tolerance developed, there would be a clockwise hysteresis when the effect was plotted as a function of plasma methadone concentration during the infusion and washout. If methadone's concentration-effect relationship was associated with a significant degree of lag (counterclockwise hysteresis), then the development of some degree of tolerance could be manifest as a shift from counterclockwise to a lesser or nonmeasurable hysteresis. This methodology enables the assessment of acute tolerance, which did not occur in this study patient.

OPIOID PEPTIDES AND TOLERANCE

The degree to which chronic opioid use confers tolerance to opioid peptides was studied in patients receiving either the peptide [smd[nm-ala-[smd[nm-leu-enkephalin (DADL) or beta-endorphin. Moulin et al. (36) demonstrated that DADL produced analgesia when administered intrathecally to patients tolerant to systemic opioids. DADL is a peptide that binds selectively to delta receptors, and it was hypothesized that, in patients chronically receiving a mu agonist drug such as morphine, DADL would produce analgesia. In this study, however, without knowing the relative potency of DADL by the intrathecal route, as compared to morphine, the issue of complete or incomplete tolerance could not be fully identified.

In a second study to assess the clinical pharmacology of human beta-endorphin in humans, where various doses of human beta-endorphin were injected intracerebro[chventricularly in a tolerant patient with chronic pain, a dissociation between tolerance to analgesic effects and to the neuroendocrine effects of beta-endorphin were observed in a tolerant patient (12). The patient under study received varying doses from 0.1 mg up to 7.5 mg of beta-endorphin. Following ICVT administration, a rise in plasma prolactin and a decrease in growth hormone occurred after doses ranging from 0.1 to 7.5 mg without behavioral and analgesic effects. At a dose of 7.5 mg analgesic effects could be demonstrated. The time and dose differences between the neuroendocrine effects in analgesic and behavioral response could be due to differences in proximity to receptors to the ventricular surface, differences in receptor affinity, differences in receptors, or the development of tolerance to the analgesic but not to the neuroendocrine effects. These findings again support the observation that the rates of tolerance development to analgesia and neuroendocrine effects vary, with analgesic tolerance developing more rapidly.

From these studies designed to address the issue of tolerance development, several general conclusions can be made: (i) tolerance develops to the analgesic effects of opioids with chronic administration, (ii) the rates to which tolerance develops to the varying effects of opioids differ, (iii) cross-tolerance is incomplete, and (iv) tolerance to one opioid drug does not confer complete tolerance to another opioid drug.

CLINICAL SURVEY DATA ON TOLERANCE

In order to address the extent to which patients chronically receiving opioids increase their opioid requirements over time, a variety of survey studies to assess the patterns of drug use in patients with cancer and pain have been undertaken (5,15,16,24,32,36,37,39,40,42,45,48,51,55). From these studies, three patterns of drug use emerge: (i) rapidly escalating doses of opioids associated with escalating pain and/or anxiety, (ii) stable doses of opioids for long periods of time (weeks to months) without dose escalation and/or reduction, and (iii) discontinuance of opioid drugs following effective relief of pain by anticancer therapies or anesthetic or neurosurgical procedures (9-11). These patterns have now been described in the Memorial Sloan-Kettering Cancer Center (MSKCC) studies in an outpatient cancer pain clinic population (24), in an inpatient Pain Service population of patients receiving continuous infusions of opioids (42), and in a population of patients with far-advanced disease followed by the Supportive Care Program of MSKCC (5) and in a group of patients with chronic nonmalignant pain (40).

Numerous other investigators have observed this phenomenon and have defined a series of patterns of opioid use. Twycross (55) early on pointed out that most patients with pain and cancer did not require escalation of their doses of opioids for pain control. More recently, in a survey of 550 cancer patients who were treated with morphine for a total of 22,525 treatment days, patients obtained pain relief using an average oral morphine dose of 82.4 mg (46). In more than 50% of the patients studied, the dose either remained stable throughout the course of therapy or was reduced. In more than 50% of dose or therapy changes, these were caused by an increased intensity of pain. Figure 10 , from Schug et al. (46), demonstrates the comparison of mean daily oral morphine doses in different patient populations studied. Moreover, in this same study of 550 patients, Figure 11 demonstrates the effect of other successful interventions for pain control on oral morphine dosage, further confirming the fact that reduction in opioid requirements can definitely occur. Table 3 lists a wide variation in the use of opioids in a population of patients followed in the MSKCC Supportive Care Program to point out the fact that there are patients who may require large amounts of opioids during the course of their illness but that the majority of patients maintain stable doses for long periods of time (5). Several other studies support these observations. Onofrio and Yaksh reviewed the long-term pain relief produced by intrathecal morphine infusion in 53 patients; Figure 12 details their observations (39). The concentration of spinal drug administered was incremented. In a study population of patients who survived in excess of 16 weeks after pump implantation, the daily infusion dose rose from 3.7;pm0.3 mg/day in the second week to 9.5;pm2.1 mg/day by week 16. The daily analgesic equivalents of systemic morphine (DAEM) rose from 0.6;pm0.1 in week 2 to 2.1;pm1.1 in week 16. Sixty-five percent of these patients were considered to have good to excellent pain control. Dose escalation occurred but patients had periods of stable opioid requirements.

Chapman and Hill (3) studied cancer patients with painful oral mucositis and compared drug use in patients self-administering morphine for 2 weeks compared to [chcontrols who received the drug via routine staff-controlled continuous infusion procedures. The pain relief was essentially identical in the two groups. The self-administering group took less morphine than the continuous infusion group. [chTolerance did not develop to the analgesic effects of the drugs. This well-studied population further supports the clinical observations that patients do not escalate their dose requirements unless there is a change in pain intensity.

Chapter 20 table 3: Opioid requirements observed in 100 cancer patients over 24 hr expressed as morphine sulfate equivalents in the MSKCC Supportive Care Program
IM morphine equivalents (mg) No. of patients
 5-99   34 
 100-199   19 
 200-299   13 
 300-699   17 
 900-1,999    4 
 2,000-5,000   10 
 7,992    1 
 19,200    1 
 35,165    1 
 5-35,165 (range)  100 (total)
MSKCC, Memorial Sloan-Kettering Cancer Center; IM, intramuscular.
Of note, similar data have been published for patients with chronic nonmalignant pain (40). Several studies describing the pattern of opioid use in this population of patients have observed that chronic low-dose therapy without significant escalation is the rule. Three chronic pain patients receiving pethidine were studied pharmaco[chkinetically (13). These patients maintained stable minimal effective concentrations of pethidine with effective analgesia for long periods of time.

These clinical observations provide a strong case for the concept that chronic opioid use does not imply continual dose escalation. Moreover, it strongly supports the concept that multiple factors such as pain intensity, type of pain, and PK and genetic factors may play a role in dose escalation. Tolerance is only one effect of dose escalation (41).

MANAGEMENT OF TOLERANCE IN PATIENTS WITH PAIN

That tolerance develops is not problematic unless tolerance to analgesia occurs before tolerance to the limiting side effects of the opioids. The first sign of tolerance in patients is the complaint that the drug works for only 3 hr instead of 4 hr. Studies in patients reveal that there is no limit to tolerance. The dose that works is the dose that works. Increasing the dose to achieve analgesia may require a two- to 10-fold increase in dose because the dose-effect relationship is based on a log dose concentration. Practically speaking, this means that doubling the dose may be necessary to produce an associated analgesic effect. Figure 13 graphically depicts the slow increase in drug requirements in a 62-year-old man with carcinoma of the colon and abdominal pain. His pain was controlled with a controlled-release morphine preparation that required increasing doses each time he developed a further complication of his disease. Each escalation was associated with effective pain relief and minimal side effects. He died with adequate pain control at a dose four times his starting dose. In a second example, Figure 14 depicts a 56-year-old woman with pancreatic cancer and a stable pain syndrome who was maintained on oral morphine and the intermittent use of aspirin for approximately 50 weeks when she developed increasing pain associated with objective signs of tumor progression. Oral doses of morphine were increased but because of gastrointestinal obstruction she was switched to intravenous morphine, first by intravenous bolus and then by continuous infusion. In the last 4 weeks of life, progressive escalation of her dose to 6,000 mg of morphine sulfate per day by continuous intravenous infusion was necessary to control her pain. She remained awake, without respiratory compromise, until her death from gram-egative sepsis. Tolerance did not present a problem while her pain syndrome was stable, but with progression of her disease and increasing pain severity, large dose requirements were needed to provide pain relief. This clinical experience suggests that there is no limit to tolerance. The patient was able to maintain analgesia while tolerant to the sedative and respiratory depressant effects.

Because cross-tolerance is incomplete, it is often useful to switch a patient from one opioid analgesic to another. This has been well described in the literature. In one study (11) assessing the outcome of 46 continuous infusions of opioids in 36 patients treated at MSKCC, pain relief was achieved in 28 infusions but was unsuccessful in 18. In six of these 18 unsuccessful infusions, persistent pain and/or intolerable side effects led to a trial of continuous infusions with an alternative opioid. Two of these replacement infusions produced analgesia not achieved by the original infusion and two yielded continuous analgesia with significantly fewer side effects.
Other ways to manage the development of tolerance are to use local anesthetics combined with opioid analgesics by the epidural route and the use of neurosurgical and/or neurolytic procedures that are associated with dramatic reduction in pain and rapid reversal of tolerance. This prominent plasticity and its rapid decrement with improved pain control by alternative methods suggests that the mechanisms that underlie tolerance must involve neuronal substraits as the prime site. The use of adjuvant drugs combined with narcotics, along with the previously described use of anesthetic, neurosurgical, and other opioid drug approaches, provide numerous methods to manage tolerance development in the patient with pain.

CURRENT THEORIES OF TOLERANCE

At the current time, the cellular and molecular mechanisms of tolerance are under intense scrutiny. With our increased understanding of the mechanisms of opiate receptor activation, both adenylate cyclase inhibition and changes in ion-channel activities may play a role in tolerance (17,29,38,47,49,53,54). Opioid receptors use G protein (guanine nucleotide regulatory protein) as a coupling component in their signal transduction mechanism. G proteins ``couple'' receptors to effector proteins and regulate both effector activity as well as the receptor affinity for hormonal ligands. There are at least five families of G proteins and each G protein has three subunits that undergo complex interaction with each other to modulate receptor and effector activities (43). It is currently believed that the inhibitory G protein (Gi) is associated with inhibition of adenylate cyclase activity by delta opioids. Kp channel activation by mu and delta receptors is probably mediated by Gi or Go. Kappa receptor activation inhibits voltage-sensitive calcium channels and, via Go, calcium currents are inhibited by delta receptor activation. Su et al. (50) described these molecular mechanisms and proposed a working model for in vivo opioid tolerance. They suggest that there are potentially two mechanisms of opioid tolerance=mopioid receptor down[chreg[chulation and uncoupling of receptors from G proteins. Morphine does not produce downregulation but does induce uncoupling of receptors. It remains unclear why nonpeptide opioids like morphine are not able to induce receptor downregulation.

In studies to further address the mechanism of tolerance, several investigators addressing the mechanisms of these neuroadaptive behaviors have focused on the concept that tolerance and physical dependence are experience-dependent, reversible changes and can be considered hallmark examples of behavioral plasticity (53,54). N -methyl-[smd[nm-aspartate (NMDA) receptors are a subclass of excitatory amino acid receptors that, once activated, produce calcium influx in neurons. From numerous studies, it has now been demonstrated that NMDA receptor antagonists, including the noncompetitive antagonist, MK801, and the nonselective excitatory amino acid antagonist, kynurenic acid, inhibit tolerance to the analgesic effects of repeated morphine administration without affecting either pain responsiveness on its own or the acute analgesic actions of morphine (33,54). Further studies have now demonstrated that NMDA antagonists not only prevent tolerance development but also can reverse it once it has occurred.

Trujillo and Akil (53) proposed the following series of events that occur with chronic opioid administration as a means to define a mechanistic hypothesis for tolerance. They suggest that, following the exogenous administration of an opiate, opiate receptors are affectively coupled to G proteins and the acute actions of the drugs are manifest.

With chronic opioid receptor occupation, a functional decoupling of opioid receptors from G proteins occurs and the acute effects of the drugs decrease. Tolerance develops and higher doses of opiate are necessary to trigger the second messenger response to produce physiologic and behavioral effects. With chronic opioid exposure, there is a decrease in endogenous opioid biosynthesis that may have no obvious consequences but is made evident when the exogenous drug is terminated. A rebound hyperexcitability of opioid responsive neurons occurs resulting from physiologic changes within opioid-responsive neurons themselves, from the decreased activity of endogenous opioid neurons, or from excessive activity of excitatory inputs. The increased firing of these neurons is what is described as the syndrome of opiate abstinence. Eventually, in the absence of any exogenous opioid drug, opioidreceptor coupling to the G protein begins to recur, as does recovery of endogenous opioid biosynthesis. In short, opioid tolerance and dependence may be related both to a functional decoupling of opioid receptors from second-messenger events and a decrease in the availability of endogenous opioid peptides.

One question that has arisen is: What is the mechanism by which an NMDA receptor antagonist mediates this impact on tolerance? The question is whether the receptors act directly at endogenous opioid synapses or at a site or sites distal to these. One hypothesis is that the NMDA receptor-mediated increase in intracellular calcium may be involved in the changes in receptor coupling, opioid peptide biosynthesis, or both (17). If this is correct, then NMDA receptor antagonists would inhibit opiate tolerance and physical dependence by directly interfering with the cellular and molecular changes thought to be involved in these phenomena.

Further evidence to support this hypothesis comes from studies with nitric oxide (26). Pasternak et al. implicated nitric oxide in the mechanisms of mu receptor tolerance and dependence. They demonstrated that the nitric oxide synthase inhibitor (NO-arginine) (N[cf11]G -itro-[sml[nm-arginine) blocks the development of tolerance to morphine in a dose-dependent manner. The actions are restricted to the mu opiate morphine. This agent did not prevent tolerance to kappa or kappa agents. These data support the observation that the development of tolerance to mu and kappa drugs involves pharmacologically distinct mechanisms of action. Moreover, it suggests that this selective effect of nitric oxide synthase inhibitors to interfere with tolerance may involve a parallel noninteracting system with the NMDA antagonists. The authors suggest that it is unlikely that the nitric oxide synthase inhibitors are interfering with learning processes because of their selective effect on tolerance to mu rather than to kappa analgesics. Of particular interest is the fact that nitric oxide synthase is an enzyme identified within specific regions of the brain known to contain opioid receptors and to be important in the production of analgesia. It does not appear at the present time that nitric oxide synthase corresponds to specific sites of mu receptors or other known opioid receptor subtypes and may therefore not play a widespread role in opioid action. However, the ability of both nitric oxide synthase inhibitors and an NMDA antagonist to reduce tolerance provides a great advantage in the use of opioid analgesics. The clinical utility of these agents will probably, however, reside in their side-effect profile. Both of these observations provide the impetus to develop clinically useful drugs that may impede the development of tolerance.

REFERENCES

  1. Adler MW. Mechanisms of opioid tolerance and dependence: symposium summary. In: Harris LS, ed. Problems of drug dependence 1986: proceedings of the 48th annual scientific meeting. Washington, DC: The Committee on Problems of Drug Dependence, Inc., NIDA Res. Monogr. 1989;76:63-68.
  2. Bruera E, MacMillan K, Hanson J, MacDonald N. The cognitive effects of the administration of narcotic analgesics in patients with cancer pain. Pain 1989;39:13-16.
  3. Chapman CR, Hill HF. Prolonged morphine self-administration and addiction liability. Cancer 1989;63:1636-1644.
  4. Cox BM. Molecular and cellular mechanisms in opioid tolerance. In: Basbaum AL, Besson J-M, eds. Towards a new pharmacotherapy of pain. New York: Wiley, 1991:137-156.
  5. Coyle N. Continuity of care for the cancer patient with chronic pain. Management of cancer pain. Cancer 1989;63(suppl):2289-2293.
  6. Dafters R, Odber J. Effects of dose, interdose interval, and drug-signal parameters on morphine analgesic tolerance: implications for current theories of tolerance. Behav Neurosci 1989;103:1082-1090.
  7. Dafters R, Odber J, Miller J. Associative and non-associative tolerance to morphine: support for a dual-process habituation model. Life Sci 1988;42:1897-1906.
  8. Fields HL (Rapporteur), Cox BM, Foley KM, et al. Group report: strategies for improving the pharmacological approaches to the maintenance of analgesia in chronic pain. In: Basbaum AI, Besson JM, eds.Towards a new pharmacotherapy of pain. Dahlem Workshop Reports. New York: Wiley, 1991:205-226.
  9. Foley KM. Pharmacologic approaches to cancer pain management. In: Fields H, ed. Advances in pain research and therapy (Fourth World Congress on Pain, IASP). New York: Raven Press, 1985:629-653.
  10. Foley KM. Controversies in cancer pain, medical perspectives. Cancer 1989;63:2257-2265.
  11. Foley KM. Clinical tolerance to opioids. In: Basbaum AI, Besson JM, eds. Towards a new pharmacotherapy of pain. Dahlem Workshop Reports. New York: Wiley, 1991:181-204.
  12. Foley KM, Kourides IA, Inturrisi CE, et al. B-Endorphin: analgesic and hormonal effects in humans. Proc Natl Acad Sci USA 1979;76:5377-5381.
  13. Glynn CJ, Mather LE. Clinical pharmacokinetics applied to patients with intractable pain: studies with pethidine. Pain 1982;13:237-246.
  14. Goudie AJ, Demellweek C. Conditioning factors in drug tolerance. In: Goldberg SR, Stolerman IP, eds. Behavioral analysis of drug dependence. Orlando: Academic Press, 1986:225-285.
  15. Goughnour BR, Arkinstall WW, Stewart JH. Analgesic response to single and multiple doses of controlled-release morphine tablets and morphine oral solution in cancer patients. Cancer 1989;63:2294-2297.
  16. Gourlay GK, Cherry DA, Cousins MJ. A comparative study of the efficacy and pharmacokinetics of oral methadone and morphine in the treatment of severe pain in patients with cancer. Pain 1986;25:297-312.
  17. Gross RA, MacDonald RL. Dynorphin A selectively reduces a large transient (N-type) calcium current of mouse dorsal root ganglion neurons in cell culture. Proc Natl Acad Sci USA 1987,84:5469.
  18. Hinson RE, Siegel S. Nonpharmacological bases of drug tolerance and dependence. J Psychosom Res 1982;26:495-503.
  19. Houde RW. The use and misuse of narcotics in the treatment of chronic pain. In: Bonica JJ, ed. Advances in neurology: international symposium on pain. New York: Raven Press, 1974:527-536.
  20. Houde RW. Nathan B. Eddy Memorial Award Lecture: the analgesic connection. In: Harris LS, ed. Problems of drug dependence 1984: proceedings of the 46th annual scientific meeting. Washington, DC: The Committee on Problems of Drug Dependence, Inc., NIDA Res. Monogr. 1985;55:4-13.
  21. Houde RW, Wallenstein SL, Beaver WT. Evaluation of analgesics in patients with cancer pain. In: Lasagna L, ed. Clinical pharmacology. Section 6, vol 1, international encyclopedia of pharmacology and therapeutics. Oxford: Pergamon, 1966:59-98.
  22. Inturrisi CE, Portenoy RK, Max MB, Colburn WA, Foley KM. Pharmacokinetic-pharmacodynamic (PK-PD) relationships of methadone infusions in patients with cancer pain. Clin Pharmacol Ther 1990;47:565-577.
  23. Kalant H. Nathan B. Eddy Memorial Award Lecture: tolerance and its significance for drug and alcohol dependence. In: Harris LS, ed. Problems of drug dependence 1986: proceedings of the 48th annual scientific meeting. Washington, DC: The Committee on Problems of Drug Dependence, Inc., NIDA Res. Monogr., 1987;76:9-19.
  24. Kanner RM, Foley KM. Patterns of narcotic drug use in a cancer pain clinic. In: Research developments in drug and alcohol use. New York: NY Academy of Science, 1981:161-172.
  25. Kantor TG. Pentazocine: clinical analgesic studies. In: Foley KM, Inturrisi CE, eds. Advances in pain research and therapy, volume 8, opioid analgesics in the management of clinical pain. New York: Raven Press, 1986:241-246.
  26. Kolesnikov YA, Pick CG, Pasternak GW. Blockade of tolerance to morphine but not kappa opioids by a nitric oxide synthase inhibitor Proc Natl Acad Sci USA =+( in press ).
  27. Kornetsky C. Early studies of tolerance to morphine. In: Harris LS, ed. Problems of drug dependence 1986: proceedings of the 48th annual scientific meeting. Washington, DC: The Committee on Problems of Drug Dependence, Inc., NIDA Res. Monogr., 1987;76:20-28.
  28. Kreek MJ. Tolerance and dependence: implications for the pharmacological treatment of addiction. In: Harris LS, ed. Problems of drug dependence 1986: proceedings of the 48th annual scientific meeting. Washington, DC: The Committee on Problems of Drug Dependence, Inc., NIDA Res. Monogr. 1987;76:53-62.
  29. Law PY, Hom DS, Loh HH. Opiate receptor down-regulation and desensitization in neuroblastoma x glioma NG108-15 hybrid cells are two separate cellular adaptation processes. Mol Pharmacol 1983;25:413-424.
  30. Light AB, Torrance EG, Karr WG, Fry EG, Wolff WA. Opium addiction. Chicago: American Medical Association (reprinted from Arch Int Med, vols 1929-30;43-44).
  31. Lyness WH, Smith FL, Heavner JE, Iacono CU, Garvin RD. Morphine self-administration in the rat during adjuvant-induced arthritis. Life Sci 1989;45:2217-2224.
  32. Macaluso C, Foley KM. Cordotomy for lumbosacral, pelvic, and lower extremity pain of malignant origin: safety and efficacy. Neurology 1988;38:110.
  33. Marek P, Ben-Eliyahu S, Gold M, Liebeskind JC. Excitatory amino acid antagonists (kynurenic acid and MK-801) attenuate the development of morphine tolerance in the rat. Brain Res 1991;547:77-81.
  34. Martin WR. Drug addiction I & II. New York: Springer, 1977.
  35. Martin WR, Jasinski DR. Physiological parameters of morphine dependence in man=mtolerance, early abstinence, protracted abstinence. J Psychiatr Res 1989;7:9-17.
  36. Moulin DE, Max MB, Kaiko RK, et al. The analgesic efficacy of intrathecal D-Ala2-D-Leu5-[chenkephalin in cancer patients with chronic pain. Pain 1985;23:213-221.
  37. Neumann PB, Henriksen H, Grosman N, Christensen CB. Plasma morphine concentrations during chronic oral administration in patients with cancer pain. Pain 1982;13:247-252.
  38. North RA, Williams JT, Surprenant A, Christie MJ. Mu and delta receptors belong to a family of receptors that are coupled to potassium channels. Proc Natl Acad Sci USA 1987;84:5487.
  39. Onofrio BM, Yaksh TL. Long term pain relief produced by intrathecal morphine infusion in 53 patients. J Neurosurg 1990;72:200-209.
  40. Portenoy RK, Foley KM. Chronic use of opioid analgesic in nonmalignant pain: report of 38 cases. Pain 1986;25:171-186.
  41. Portenoy RK, Foley KM, Inturrisi CE. The nature of opioid responsiveness and its implications for neuropathic pain: new hypotheses derived from studies of opioid infusions. Pain 1990;43:273-286.
  42. Portenoy RK, Moulin DE, Rogers AG, Inturrisi CE, Foley KM. Continuous intravenous infusion of opioids in cancer pain: review of 46 cases and guidelines for use. Cancer Treat Rep 1986;7:575-581.
  43. Rosenthal W, Hescheler J, Trautwein W, Schultz G. Control of voltage-dependent Ca2p channels by G protein-coupled receptors. FASEB J 1988;2:2784.
  44. Rosow C. Acute and chronic tolerance relevance for clinical practice. In: Harris LS, ed. Problems of drug dependence 1986: Proceedings of the 48th annual scientific meeting. Washington, DC: The Committee on Problems of Drug Dependence, Inc., NIDA Res. Monog., 1987;76:29-34.
  45. Schug SA, Zech D, Doerr U. Cancer pain management according to WHO analgesic guidelines. J Pain Symptom Management 1990;5:27-32.
  46. Schug SA, Zech D, Grond S, Jung H, Meuser TH, Stobbe B. A long-term survey of morphine in cancer pain patients. J Pain Symptom Management 1992;7:259-266.
  47. Sharma SK, Klee WA, Nirenberg M. Dual regulation of adenylate cyclase accounts for narcotic dependence and tolerance. Proc Natl Acad Sci USA 1975;72:3092.
  48. Sjorgren P, Banning A. Pain, sedation and reaction time during long-term treatment of cancer patients with oral and epidural opioids. Pain 1989;39:5-11.
  49. Smith AP, Law P-Y, Loh HH. Role of opioid receptors in narcotic tolerance/dependence. In: Pasternak GW, ed. The opiate receptors. Clifton, NJ: Humana Press, 1988:441-489.
  50. Su YF, Wong CS, Russell RD, Nishino K, Watkins WD, Chang KJ. Mechanisms of opioid tolerance. In: Short CE, Van Poznak A, eds. Animal pain. New York: Churchill Livingstone, 1990:259-268.
  51. Thirlwell MP, Sloan PA, Maroun JA, et al. Pharmacokinetics and clinical efficacy of oral morphine solution and controlled-release morphine tablets in cancer patients. Cancer 1989;63:2275-2283.
  52. Tress KH, El-Sobky AA, Aherne W, Piall E. Degree of tolerance and the relationship between plasma morphine concentration and pupil diameter following intravenous heroin in man. Br J Clin Pharmacol 1978;5:299-303.
  53. Trujillo KA, Akil H. Opiate tolerance and dependence: recent findings and synthesis. New Biologist 1991;915-923.
  54. Trujillo KA, Akil H. Inhibition of morphine tolerance and dependence by the NMDA receptor antagonist MK-801. Science 1991;51:85-87.
  55. Twycross RG. Clinical experience with diamorphine in advanced malignant disease. Int J Clin Pharmacol Ther Toxicol 1974;9:184-198.
  56. Vaught JL. What is the relative contribution of mu, delta, and kappa opioid receptors to antinociception and is there cross-tolerance? In: Basbaum AL, Besson J-M, eds. Towards a new pharmacotherapy of pain. New York: Wiley, 1991:121-136.
  57. Wikler A. Recent progress in research on the neurophysiologic basis of morphine addiction. Am J Psychiatry 1948;105:329:338.
  58. Wuster M, Costa T, Gramsch C. Uncoupling of receptors is essential for opiate-induced desensitization (tolerance) in neuroblastoma x glioma hybrid NG108-15 cells. Life Sci 1983;33:341-344.
  59. Yaksh TL. Tolerance: factors involved in changes in the dose-effect relationship with chronic drug exposure. In: Basbaum AI, Besson J-M, eds. Towards a new pharmacotherapy of pain. Dahlem Workshop Reports. New York: Wiley, 1991:157-180.