The use of opioids for chronic noncancer pain (e.g., neuropathic pain, mus- culoskeletal pain) is controversial. Despite the fact that many patients have received opioids chronically (Portenoy et al. 1991), we do not really know how well they work. Opioid efﬁcacy in musculoskeletal pain and neuropathic pain has been claimed in a number of case reports and uncontrolled open studies. Few controlled studies investigating the efﬁcacy and side effects of opioids in these clinical settings are available (Caldwell et al. 1999; Moulin et al. 1996; Peloso et al. 2000; Watson and Babul 1998). At the moment, the maximum duration of opioid treatment investigated in a double-blind placebo-controlled study is 9 weeks (Moulin et al. 1996). Most trials found a reduction in subjective pain scores, but only one study examined in de- tail much more important parameters such as psychosocial features, quality of life, drug dependence, and functional status (Moulin et al. 1996). No sig- niﬁcant improvements in any of the latter parameters were detected, and there was a lack of overall patient preference for the opioid. Most authors concluded that morphine may confer analgesic beneﬁt with a low risk of addiction, but that it is unlikely to yield psychological or functional improve- ment (Watson and Babul 1998). Adverse opioid side effects were reported in all of these investigations and led to the drop-out of large numbers (up to 60%) of patients. Thus, there is a lack of prospective controlled studies ex- amining the long-term (at least several months) administration of opioids. Future studies need to demonstrate positive outcomes, not only in subjec- tive pain reports, but, more importantly, also in terms of reduced depression, functional improvement, reemployment and decreased use of the healthcare system.
Many recent studies have focused on the activation of opioid receptors outside the CNS. One of the most extensively studied and most successful ap- plications is the intraarticular injection of morphine (Kalso 2002). In dental surgery, peripheral antinociception was detected after local morphine appli- cation (Likar et al. 1998). In patients with chronic arthritis local morphine has also been shown to reduce pain and possibly inﬂammation (Likar et al. 1997; Stein et al. 1999). The topical application of opioids produces antihyperalgesic effects when applied to painful ulcers and skin lesions, after burn injuries, and in cutaneous pain (Kolesnikov et al. 2000; Krajnik et al. 1999; Long et al.
2001; Twillman et al. 1999). The local application of morphine in patients with corneal abrasion also showed analgesic effects (Peyman et al. 1994). Novel pe- ripherally restricted κ-agonists have been investigated in humans with chronic painful pancreatitis (Eisenach et al. 2003). In addition, some studies have sug-gested a relatively reduced development of opioid tolerance in inﬂamed tissue (Stein et al. 1996).
Almost all opioids are rapidly absorbed after oral administration and many undergo substantial ﬁrst-pass metabolism in the liver. Much effort has been directed at delaying absorption with sustained-release opioid preparations because adequate duration of analgesia is a bigger clinical problem than slow onset. Sometimes almost no absorption is desirable: For example, lop- eramide and diphenoxylate are potent opioids used in the treatment of di- arrhea (and recently for topical application) because they are poorly ab- sorbed and produce minimal effects in the CNS. Opioid drugs vary tremen- dously in their lipid solubility, and this largely determines the efﬁciency of absorption from peripheral sites. For example, after sublingual administra- tion, the bioavailability of morphine is only 12%, while highly lipid solu- ble drugs like fentanyl and buprenorphine are 60%–70% absorbed into the bloodstream (Gong and Middleton 1992; Weinberg et al. 1988). Lipophilic opioids are now commercially available in buccal, intranasal, and transder- mal preparations. Using a special nebulizer apparatus, it is possible to ad- minister morphine by inhalation and achieve 63% bioavailability (Dershwitz et al. 2000).
All opioids are rapidly and extensively distributed throughout the body. Af- ter a bolus intravenous injection, the rapid increase and decrease in plasma
concentration looks quite similar for morphine and fentanyl. The difference in onset and duration of effect for these two drugs is a function of the rate at which plasma concentrations equilibrate with those in brain, at least in situations where activation of central opioid receptors is the predominant mechanism of action (i.e., without major peripheral tissue injury). CNS con- centration of a lipophilic opioid like fentanyl closely follows the concentration in plasma, rising then falling rapidly as the drug is redistributed from highly perfused tissues into muscle and fat (Hug and Murphy 1979; Lotsch 2005b). Its short duration is due to the rapidity of physical translocation out of the CNS. In comparison, the slow onset and offset of morphine are because it enters and exits the CNS slowly. Brain concentrations of morphine lag far behind those in plasma. When an opioid is administered intrathecally or epidurally, the onset of effect is determined by the rate at which the drug penetrates to reach opioid receptors in the dorsal horn. Since little metabolism occurs in the CNS, the effects are terminated by redistribution into blood vessels. Re- cent animal studies into the relative contribution of peripheral versus central opioid receptors indicate that, depending on the presence and extent of tis- sue injury and on the types of opioid receptors involved, up to 80% of an analgesic effect following systemic opioid administration may be mediated by peripheral opioid receptors (D. Labuz, S.A. Mousa, M. Schäfer, C. Stein, H. Machelska, submitted). Similar observations were made in human studies us- ing the peripherally restricted agonist morphine-6-glucuronide (Tegeder et al.
2003; Hanna et al. 2005).
Metabolism and Excretion
All clinically available opioids undergo extensive hepatic metabolism to polar metabolites that are excreted by the kidney (Lotsch 2005a). The notable ex- ception is remifentanil, which is rapidly hydrolyzed by nonspeciﬁc esterases in peripheral tissues and plasma (Servin 2003). In general, morphine and its close congeners mainly undergo synthetic biotransformation to glucuronides, while meperidine and the fentanyl derivatives undergo oxidative metabolism by cytochrome P450 enzymes. Both morphine and fentanyl have high hep- atic extraction ratios (0.7 and 0.6, respectively). This means that the clearance of these drugs is sensitive to factors that alter liver blood ﬂow. On the other hand, clearance is relatively unaffected by inducers or inhibitors of liver en- zymes. The major metabolite of morphine is the 3-glucuronide, but about 15% forms morphine 6-glucuronide (M6G), a compound that has substantial opioid agonist activity (Kilpatrick and Smith 2005). It is likely that this highly
polar metabolite does not easily enter the CNS (Portenoy et al. 1991) but may produce analgesia via peripheral opioid receptors (Hanna et al. 2005; Tegeder et al. 2003).
Acute Opioid Application
High doses of morphine, fentanyl, sufentanil, remifentanil, and alfentanil are associated with a vagus-mediated bradycardia. Severe bradycardia or even asystole is possible, especially in conjunction with the vagal stimulating ef- fects of intubation (Bowdle 1998). With the exception of meperidine, opioids do not depress cardiac contractility. Arterial blood pressure often falls as a result of bradycardia and decreased sympathetic reﬂexes. However, opi- oids have depressant effects in patients with congestive heart failure (Liang et al. 1987) and in myocardial ischemia/reperfusion-induced arrhythmias (Lee
1992). Recent evidence has also implicated opioids as having cardioprotec- tive effects (“ischemic preconditioning”), in that tissue damage after brief
periods of coronary artery occlusion was prevented by opioids (Schultz et al.
1997). Suggested mechanisms include the activation of mitochondrial ATP- sensitive K channels (KATP channels) in cardiomyocytes. Opening of mito- chondrial KATP channels is thought to be protective by preservation of mi- tochondrial integrity, by dissipation of mitochondrial membrane potential, and consequent reduction of calcium overload and apoptotic cell death (Cao et al. 2005; Cao et al. 2003). In isolated cardiomyocytes as well as sympa- thectomized intact rat hearts, it was suggested that opioid receptors func-
tionally and physically crosstalk with β-adrenergic receptors, including het- erodimerization of these receptors, counterbalance of functionally opposing
G protein signaling, and interference with downstream signaling events. As a result, the β-adrenergic receptor-mediated positive inotropic effect and an increase in cAMP are markedly attenuated after opioid receptor activation (Pepe et al. 2004).
Respiratory depression is the adverse effect most feared by clinicians. These effects are mediated by μ- and δ-receptors through the direct inhibition of rhythm-generating respiratory neurons in the pre-Boetzinger complex (PBC) of the brainstem (Manzke et al. 2003). Opioids produce a dose-dependent de-
pression of the ventilatory response to hypercarbia and hypoxia (Ladd et al.
2005; Weil et al. 1975). Respiratory depressant effects may be detectable well before clinically apparent changes in respiratory rate and depth. With increas- ing opioid doses, respiratory rate slows and tidal volume initially increases
and then eventually decreases. High doses can produce apnea, and a small number of patients still die each year from opioid-induced respiratory depres- sion. The potential is greatest in heavily co-medicated (e.g., with sedatives) patients who are unstimulated and unmonitored. In a classic study, Forrest and Belleville showed that natural sleep greatly increases the respiratory de- pressant effects of morphine (Forrest and Bellville 1964). Fentanyl, alfentanil, and sufentanil have all been reported to produce “recurrent” respiratory de- pression in postoperative patients who initially seemed to be breathing well (Becker et al. 1976). It seems likely that these episodes were actually due to variations in the level of postoperative (e.g., nociceptive) stimulation. Recently it was shown that serotonin 4(a) [5-HT4(a)] receptors are expressed in respira- tory PBC neurons, their selective activation protects spontaneous respiratory activity, and 5-HT(4a) receptors and μ-receptors affect the intracellular con- centration of cyclic AMP in opposite ways (Manzke et al. 2003). Treatment of rats with a 5-HT4 receptor-speciﬁc agonist prevented fentanyl-induced respi- ratory depression without loss of fentanyl’s analgesic activity (Manzke et al.
2003). These ﬁndings might stimulate novel therapeutic developments in the future (Eilers and Schumacher 2004).
Sedation is a frequent and serious side effect of opioid analgesics, some- times reported as fatigue or tiredness by patients (Shaiova 2005). Importantly, small subanalgesic doses of opioids can multiply the sedative potency of mi- dazolam (Kissin et al. 1990). Similar interactions have been demonstrated with propofol (Short et al. 1992) and barbiturates. Thus, for the clinician it is important to remember that patients receiving multiple sedative medica- tions simultaneously are at increased risk and need to be monitored ade- quately.
Nausea and Vomiting
Opioids stimulate nausea and vomiting by a direct effect on the chemore- ceptor trigger zone in the area postrema in the brainstem (Apfel et al. 2004; Wang and Glaviano 1954). This effect is increased by labyrinthine input, so patients who are moving are much more likely to be nauseated than those lying quietly. Prophylactic antiemetic interventions include the avoidance of other emetogenic drugs (e.g., inhalational anesthetics) by total intravenous anesthesia (e.g., with propofol). Serotonin antagonists, dexamethasone, and droperidol can be used to treat opioid-induced nausea and vomiting (Apfel et al. 2004).
Opioids depress cough by direct effects on medullary cough centers (e.g., raphe nuclei; Chou and Wang 1975; Schug et al. 1992). The structure-activity requirements for the antitussive effects are not the same as those for typi- cal μ-receptor-mediated analgesic effects. The greatest activity is seen with drugs like codeine and heroin—morphine congeners with bulky substitu- tions at the 3 position. The stereospeciﬁcity of the response is different as well: cough suppression is produced by dextroisomers of opioids (e.g., dex- tromethorphan) that do not have analgesic activity (Chung and Chang 2002; Lal et al. 1986).
The miotic effect of opioids occurs through a direct action on the autonomic (Edinger–Westphal) nucleus of the oculomotor nerve to increase parasympa- thetic tone. This effect can be detected after extremely small (subanalgesic) doses. Relatively little tolerance occurs to this effect, so even patients taking very high doses of opioids for extended periods of time (e.g., for chronic cancer pain) will continue to have constricted pupils. Pupil responses have proved use- ful for simultaneous pharmacokinetic–pharmacodynamic modeling of opioids (Zacny 2005).
Skeletal Muscle Rigidity
This phenomenon, often incorrectly termed “truncal” or “chest wall” rigidity, is actually a generalized hypertonus of striated muscle throughout the body (Benthuysen et al. 1986). It is usually seen when potent opioids are administered rapidly (e.g., during induction of anesthesia) and is most commonly produced by fentanyl and its congeners (Bowdle 1998). The mechanism appears to be an
inhibition of striatal γ-aminobutyric acid (GABA) release and an increase in dopamine production (Costall et al. 1978). Selective antagonism at the nucleus
raphe magnus in the rat can completely prevent the increase in muscle tone
(Weinger et al. 1991). During induction of anesthesia, opioid-induced muscle rigidity can render a patient difﬁcult to ventilate. Some authors have claimed that the problem does not seem to be loss of chest wall compliance but rather hypertonus of the pharyngeal and laryngeal musculature leading to narrowing of the laryngeal inlet (Arandia and Patil 1987; Bowdle and Rooke 1994). When rigidity is recognized, it must be treated with a muscle relaxant or reversed with naloxone.