Leading Article
Drugs 35: 597-603 (1988) 0012-6667/ 88/0006-05 97/$03 .50/0 © ADIS Press Limited All rights reserved.
Spinal Opioid Analgesia A Critical Update
Lars L. Gustafsson and Zsuzsanna Wiesenfeld-Hallin Departments of Clinical Pharmacology and Neurophysiology, Karolinska Institute , Huddinge University Hospital, Huddinge
Summary
S mall spinal (intrathecal or extrathecal) doses of opioids induce a long-lasting and regional analgesic effect in various experimen tal animal models. Nowadays extrathecal morphine adm inistration is considered an established method ofcontrolling postoperative and cancer-induced pain conditions. The potency of morphin e applied by the spinal route is higher than when the drug is applied by the intravenous (IV) route. Opioids which are more lipophilic than morphine will provide a marginally better analgesic effect when adm inistered by the spinal route as compared with the I V route. Se veral controlled clinical trials in postoperative patients have demonstrated that a single dose of morphine administered by the spinal route gives a more long-lasting action than a sim ilar I V dose. It is not known whether frequent patient-controlled adm inistration of morphine may provide equally good analgesia without additional side effects. The use ofspinal morphine in the treatment ofcancer-related pain is based on clinical experience only. There are risks in replacing opioid administration by the oral or I V route with spinal opioids. Mo rphine should only be used in selected cases until the advantage ofspinalopioid analgesia to control postoperative and cancer pain has been clearly defin ed in well-designed clinical studies. Spinal m orphine dosages must be individualised according to the intensity ofthe nociceptive stimuli and should take into account intra-individual variability in drug responses due to pharma cokin etic and pharma codynamic factors
Administration of various opioids by the spinal (intrathecal or extradural) route was introduced clinically in 1979, only 3 years after the demonstration of a potent analgesic effect of intrathecal morphine in rats (Yaksh & Rud y 1976). This technique attracted man y clinicians since the experimental work done in rats showed a long-lasting dose-dependent effect which was found to be naloxone reversible , regional and with no apparent effects on motor function (Yaksh & Rudy 1976). It was hoped that small doses of morphine administered by the spinal route would be devoid of side
effects and provide an analgesic action as potent as seen with local anaesthetic blocks, but without depression of motor function . Within a few years of mostly uncontrolled clinical use of spinal opioids , however, unpredicted side effects including late ventilatory depression , itching and urinary retention were reported (Bromage et al. 1982; Yaksh 1981). Today extradural 'morphine administration is used worldwide in the clinical control of pain and accounts in Sweden for as much as 25% of all extradural blocks (Rawal et al. 1987). During the past few years a number of
Spinal Opioid Analgesia
controlled clinical studies of the effect of spinal morphine administration in the postoperative period have been performed and basic knowledge about its mechanism of action has increased. It may therefore be time to critically discuss the role of spinal -opioids in the treatment of various pain conditions.
1. Basic Concepts: Spinal Modulation of Nociceptive Stimuli Endogenous opioid peptides believed to be of importance in spinal analgesia may be formed from 3 precursor molecules: proenkephalin A (PEA), proenkephalin B (PEB) and proopiomelanocortin (POM) [see Terenius 1985]. Methorphamide is the rapidly degraded natural ligand of the wreceptor and is derived from PEA. Enkephalin , also derived from PEA, binds primarily to the 0- but also to the ,a-receptor. Dynorphin is the ligand derived from PEB and is quite selective in binding to the x-receptor. {3-Endorphin, derived from POM, binds preferentially to the e-receptor. The various spinal opioid receptors modulate reaction to experimental nociceptive stimuli in animals with some selectivity (Millan 1986; Przewlocki et al. 1983; Schmauss and Yaksh 1984). The wreceptor agonists, of which morphine is the most important, are effective against noxious mechanical, thermal or chemical stimuli . The selective 0receptor agonist D-Ala 2-D-Leu 5-enkephalin (DADL) causes antinociception at the spinal level against thermal stimuli . The role of this specific 0receptor agonist in suppressing noxious mechanical and chemical inputs is unclear. Dynorphin has a weak action against noxious thermal stimuli . There is also evidence that the x-receptors can mediate antinociception against non-thermal inputs, such as visceral pain and noxious pressure, while {3-endorphin has strong anti nociceptive effect at spinal level against both thermal and non-thermal stimuli. As it is unclear whether the e-receptor exists in the spinal cord, it is possible that the effect of {3-endorphin is mediated by w or o-receptors. In physiological experiments in animals it has been shown that systemic or intrathecal morphine
a
598
suppresses input to spinal cord multireceptive cells from unmyelinated C-afferents (the majority of these are nociceptors) more than from myelinated ones (Le Bars et al. 1976). In humans evidence has also been presented that morphine selectively suppresses painful input from C-polymodal nociceptors, which respond to noxious mechanical, thermal and chemical stimuli. Morphine does not affect input from large or small myelinated afferents or from C-afferents that respond to innocuous warming (Copper et al. 1986). With increased doses of morphine the selectivity in effect may disappear since a scratching behaviour is induced by nonopioid mechanisms in rats (Yaksh et al. 1986a). In summary, there are abundant experimental data which indicate that various spinal opioid receptors can mediate the analgesic effect of spinally applied opioids. The experimental results show that the efficacyof spinal opioids in humans should vary according to the clinical type of pain.
2. Mechanism ofAction: Animal and Human Studies It is quite evident that the analgesic efficacy of spinal opioids varies according to the type of clinical pain. The effects of extradural morphine were examined in a group of cancer patients with somatic, visceral, neurogenic or cutaneous pain (Arner & Arner 1985). The pain was also classified as continuous or intermittent. It was found that continuous pain originating from deep somatic tissues could always be controlled by morphine. The continuous visceral pain was also quite responsive to morphine whereas intermittent somatic and visceral pain were affected only to a very small degree. Neurogenic and cutaneous pain were quite unresponsive. These results are in line with a variable duration of effect in postoperative pain (Stenseth et al. 1985; Weddel & Ritter 1981). For example, using an extradural dose of morphine 4 to 6mg, analgesia lasted for a minimum of 12 hours in 70, 50 or 20% of patients undergoing hip replacement (n = 252), abdominal surgery (n = 539) or thoracotomy (n = 56), respectively (Stenseth et
Spinal Opioid Analgesia
al. 1985). In the first group of patients continuous somatic type of pain is predominant. Dose-dependent antinociceptive effects of several wagonists such as lofentanil , morphine, sufentani!, fentanyl , alfentanil , methadone and pethidine (meperidine) have been demonstrated after intrathecal administration in rats and for some of the drugs in primates (Yaksh & Reddy 1981 ; Yaksh & Rudy 1977; Yaksh et al. 1986b). Partial agonists to ~- and/or x-receptors such as buprenorphine and pentazocine have shown less impressive effects (Bryant et al. 1983). Furthermore, of several tested drugs given in equipotent doses, long-lasting analgesia has so far been seen only with the hydrophilic drug morphine and with a single lipoph ilic drug, lofentanil (Yaksh et al. I986b). It should be remembered, however, that lipophilic compounds such as alfentanil, sufentanil and lofentanil produce similar analgesic effects experimentally whether a dose is given by the intravenous route or by the more complicated and hazardous extradural route (Colpaert et al. 1986; Durant & Yaksh 1986). Several opioids, such as buprenorphine, diamorphine (diacetylmorphine; heroin), fentanyl, hydromorphone, methadone, pentazocine, pethidine and alfentanil , have been given to humans by the spinal route to test their efficacy, despite the fact that experimental studies have shown a long-lasting effect only with morphine and lofentanil. Most of the studies have been uncontrolled and performed in a limited number of patients, which may explain the conflicting results about the degree of analgesic effect and the duration of analgesia in various studies. To date only morphine administered extradurally has shown a documented effect in controlling postoperative pain in a number of well-designed clinical studies (Gustafsson 1985; Rawal et al. 1982; Weddel & Ritter 1981). Thus, in Sweden the use of extradural administration of morphine only has been accepted , since 96%, 3% and I% of all patients treated with analgesics via this route were given morphine, pethidine and buprenorphine, respectively (Rawal et al. 1987). However, the number of controlled clinical studies evaluating the efficacy of intrathecal morphine
599
Lumbar
Thoracic
Lumbar
Thoracic
Epidural space CSF Spinal cord Body
Excreted Morphine
Pethidine
Fig. 1. Distribution of extradural op ioids 60 minutes after injection in humans. Less than 4% of the administered dose is pre sent in CSF or spinal cord when either pethidine or morphine is given . Higher amounts of morphine than pethidine are found in the spinal cord , both at lumbar and thoracic levels. The figure shows calculated results based on published CSF concentrations in humans (n = 12) (Sjostrom et al. 1988) and on opioid concentrations at various levels of the spinal cord after intrathecal administration in rats (Gustafsson & Post 1986).
administration in the postoperative period is limited (Bengtsson et al. 1983).
3. Pharmacokinetics Data on the disposition of spinal opioids in humans are incomplete, since medical risks exclude sampling of cerebrospinal fluid (CSF) from rostral levels in most cases. Today there are as yet no reliable non-invasive techniques which can be used to quantify distribution processes into the spinal cord. Transfer to the body by uptake into the vascular system is the major route of distribution of extradural opioids, irrespective of their physicochemical properties (fig. I). In fact, after extradural administration morph ine plasma concentrations are at least as high as after intramuscular injection (Gustafsson 1985; Nordberg et al. 1983). Maximal morphine concentrations are achieved within 10 to 30 minutes after injection . At most, I to 2% of the extradural dose is transferred into CSF (Nordberg 1986) which implies that the intrathecal dose has to be kept low as compared with the extrathecal to avoid concentration-dependent side effects such as
600
Spinal Opioid Analgesia
ventilatory depression, itching and urinary retention (Rawal et aI. 1987). No human data exist on how the physicochemical properties of opioids affect the rate of diffusion into the spinal cord. Since analgesic effects of opioids in man seem to agree with experimental results in animals, the data on the distribution of opioids within the spinal canal in animals may also be valid in humans. The lipophilicity of a drug seems to inversely correlate to the fraction of the dose which is recovered in the spinal cord after spinal administration in animals. Lipophilic opioids diffuse easily into the spinal cord and such a drug will be rapidly distributed to the systemic circulation through the venous system of the spinal cord. Hydrophilic opioids, on the other hand, diffuse slowly into the venous system draining the spinal cord. Only 7% of an intrathecal dose of pethidine (lipophilic), but as much as 27% of morphine (hydrophilic) , was recovered in the neuroaxis of the rat 15 minutes after injection (Gustafsson et aI. 1985). These results agree with the short-lived effect of pethidine and the long-lasting analgesia seen with morphine (Gustafsson & Post 1986). However, lofentanil shows long-lasting analgesia in animals despite its extreme degree of lipophilicity which may be explained by its pronounced receptor affinity (Durant & Yaksh 1986). The short-lived analgesic effect for lipophilic opioids has previously been shown when several opioids were injected intraventricularly in experimental models (Herz & Teschemacher 1971). Only a small fraction of the injected dose was recovered in brain of lipophilic opioids whereas much higher amounts of hydrophilic drugs were recovered . However, in future new types of opioids with an extreme degree of receptor affinity may be synthesised so that all opioid receptors will be occupied despite a pronounced rate of distribution from the spinal cord into the systemic Circulation. For hydrophilic compounds such as morphine the ratio of CSF concentration at site of injection versus plasma concentrations is higher with intrathecal than with extradural administration (Nordberg et aI. 1983, 1984). Ratios of 100-200: I and 1000-5000: I have been seen with extradural
and intrathecal administration, respectively, which is explained by a slow rate of systemic uptake after direct morphine application into the intrathecal space. Some pharmacokinetic differences between intrathecal and extradural drug administration may be relevant to the mechanism of action. Negligible morph ine plasma concentrations are detected after intrathecal administration of doses less than Img. A supraspinal site of the analgesic effect therefore seems unlikely. In contrast, supraspinal factors may significantly contribute to the effect of extradural morphine since the plasma concentrations of morphine administered in this fashion are similar to those seen after intramuscular administration. Unless a significant part of a morphine dose applied by the intrathecal route is distributed supraspinally with CSF, analgesia should be more regional than with extradural administration. The latter route gives a morphine analgesia which is induced by both spinal and supraspinal mechanisms since a substantial part of the dose is systemically distributed within a short time period (Yeung & Rudy 1980).
4. Adverse Side Effects Adverse side effects of spinal opioid analgesia include ventilatory depression, itching and urinary retention (Bromage et al. 1982; Cousins & Mather 1984; Gustafsson 1985; Gustafsson et al. 1982). During the first few years of clinical use the incidence of ventilatory depression requiring naloxone was calculated to be 0.25 to 0.4% and 4 to 7%after extradural and intrathecal morphine administration , respectively (Rawal & Wattwil 1984). With increased knowledge of risk factors (high age, thoracic dose administration, reduced ventilatory capacity, concomitant systemic administration) the incidence of delayed ventilatory depression was found to be as low as 0.09% in a recent survey (Rawal et al. 1987). Importantly, this effect on ventilation is dose dependent. Ventilatory depression is insidious as it may be seen as late as 12 hours postinjection because of slow rostral CSF-transport of morphine (Gustafsson 1985). It may take 2 to
Spinal Opioid Analgesia
3 hours for a compound injected at the lumbar level to reach the cisterna magna and then still some additional hours to reach the basal cisterns. With an extradural dose of morphine 10mg the ventilatory depression was maximal at 5 hours as evaluated by the effects on the PaCOrresponse curve (an experimental model which can be used to quantify the degree of ventilatory depression), but the effect could still be detected at 24 hours (Rawal & Wattwil 1984). Urinary retention is seen in 20 to 40%of patients after the administration of 4 to 6mg extradural morphine for postoperative pain (Cousins & Mather 1984; Rawal et al. 1983). However, decreased sensory urge to empty the urinary bladder and reduced detrusor activity are present with doses as low as 2mg (Rawal et al. 1983). This effect is naloxone-reversible . During long term morphine administration by the spinal route, urinary retention is seldom a clinical problem. Itching can be either regional or generalised. It may even be present on the nose because of rostral transport in the CSF. Doses of 4 to 6mg morphine produce an incidence of itching of 5 to 10% and this side effect may also be relieved by naloxone administration (Cousins & Mather 1984).
5. Unknown Facts Several controlled clinical trials have shown that in the postoperative period I to 3 doses of extradural morphine (usually 2 to 6mg each) provide better analgesia as evaluated by the visual analogue scale than the usual clinical practice of administration of opioids by the systemic route (see Cousins & Mather 1984; Gustafsson 1985; Lanz et al. 1982; Weddel & Ritter 1981). However, we are not aware of any well-designed study comparing the effect of single or repeated spinal doses (extradural or intrathecal) with frequent-dose administration of opioids by the intravenous route (e.g. with patient-controlled analgesic pumps). With such repeated-dose administration the analgesic effect could be as good as with spinal dose administration since the lower potency of opioids after sys-
601
temic as compared to spinal administration will be taken into account. The hourly dose of morphine has been suggested to be 5 times higher when given by the intravenous as compared to the extradural route (Sjostrom et al. 1987). Until such comparative studies have been carried out it is not possible to answer the questions of whether the therapeutic effect and/or the incidence of side effects are different between morphine administration by the 2 routes (given in equipotent doses). Such studies are essential, as systemic opioid administration by patient-controlled pumps poses no practical problems in postoperative units, making complicated and potentially hazardous administration by the extradural route of doubtful value. Theoretical knowledge speaks, however, in favour of an analgesic efficacy of extradural morphine that is as good as that after intravenous administration, but with a lower incidence of systemic side effects, since plasma morphine concentrations are lower after repeated morphine administration by the extradural route than after administration by the intravenous route (Sjostrom et al. 1987). Several recent uncontrolled studies have reported that extradural morphine administration may be used to treat chronic cancer pain (Banning et al. 1986; Carl et al. 1986; Malone et al. 1985). However, randomised studies comparing the efficacy of oral versus spinal opioid administration in various chronic pain conditions need to be performed, since in most cases individualised oral morphine treatment controls cancer pain as well (Sawe 1984). Until results from such studies are available, morphine given by the spinal route should be considered a trial and error method . Clinical experience suggests that patients who are treated with oral morphine and who experience pronounced adverse effects such as constipation, dizziness, nausea and headache prefer to be treated with spinal morphine. As stated above, with this route a low daily morphine dose can be given and lower morphine plasma concentrations achieved. A potential area of interest is a combined treatment with spinal opioids (especially morphine) and local anaesthetic drugs, which have an additive ef-
Spinal Opioid Analgesia
feet. A prolonged duration of analgesia may thereby be achieved with a single injection in the early postoperative period. Such a drug combination by the spinal route has to be explored in clinical studies together with other spinal drug combinations, including opioids + o-agonists and opioids + serotonergic drugs (Tamsen & Gordh 1985; Yaksh & Reddy 1981).
6. Present Role of Spinal Opioids in the Management of Pain A single spinal dose of morphine has, in several controlled clinical studies, shown a more long-lasting effect than a dose given by the intravenous route in the postoperative period. As an example, a dose of morphine (2 to 4mg) administered by the extradural route and 8 to 12mg as intravenous injection gives a duration of effect of 12 to 16 hours, and 4 to 6 hours, respectively, after hip replacement (Cousins & Mather 1984; Gustafsson 1985; Yaksh & Reddy 1981). However, it is unknown if frequent patient-controlled administration by the systemic route may provide equally effective analgesia without additional side effects. Therefore, morphine administration by the spinal route still has to be considered as an analgesic technique which should be used in special patients only. Spinal morphine dosages must be individualised according to the type of pain a patient experiences. Intraindividual variability in the pharmacokinetics and the pharmacodynamics of the drug should also be considered. Spinal morphine has no value in controlling obstetric pain (Cousins & Mather 1984; Gustafsson 1985) while its use to control cancer pain should only be considered in patients who are expected to experience side effects from oral morphine because of high morphine plasma concentrations. Clearly, there is a need for comparative studies of the value of long term spinal morphine administration. Since lipophilic opioids are rapidly distributed from the spinal cord, short-lived effects are expected, unless an extremely high receptor affinity of a drug compensates for this rapid distribution from the spinal cord. Lofentanil is the single lipo-
602
philic opioid tested which shows long-lasting analgesic effects in animals and man. Its use by the spinal route is, however, doubtful since similar duration of effects has been observed whether the same dose is given by the intravenous or the extradural route in rats (Durant & Yaksh 1986). Human studies which conclude that lipophilic Itopioids given by the spinal route provide a potent and long-lasting analgesic effect should therefore be interpreted with caution.
Acknowledgements This work was supported by grants from the Swedish Medical Research Council (3902 and 7913), the Swedish Cancer Fund, the Stockholm County Council and research funds of the Karolinska Institute . Expert secretarial assistance was provided by Karen McManus .
References Arner S. Arner B. Differential effects of epidural morph ine in the treatment of cancer-related pain . Acta Anaesthesiologica Scandinavica 29: 32-36. 1985 Banning AM. Schmidt JF . Chraemmer-Jergensen B. Risbo A. Comparison of oral cont rolled-release morphine and epidural morphine in the management of postoperative pain. Anesthesia and Analgesia 65: 385-388. 1986 . Bengtsson M. LOfstrom B. Merits H. Postoperative pain relief with intrathecal morphine after major hip surgery. Regional Anaesthesi e 8: 139-143. 1983 Brornage PRo Camporesi EM. Durant PAC, Nielsen CH. Nonrespiratory side effects of epidural morphine. Anesthesi a and Analgesia 61: 490-495. 1982 Bryant RM. Olley JE. Tyers MB. Antinociceptive actions ofmorphine and buprenorphine given intrathecally in the conscious rat. British Journal of Pharmacology 78: 659-663. 1983 Carl P. Crawford ME. Raulo O. Bach V. Long term treatment with epidural opioids, Anaesthesia 41: 32-38. 1986 Colpaert FC. Leysen JE. Michiels M. van den Hoogen RHWM . Epidural and intravenous sufentan il in the rat: analgesia. opiate receptor binding. and drug concentrations in plasma and brain . Anesth esiology 65: 41-49. 1986 Copper BY. Vierck Jr CJ. Yoemans DC. Selective reduction of second pain sensations by systemic morph ine in man . Pain 24: 93-116. 1986 Cousins MJ. Mather LE. Intrathecal and epidural administration of opioids. Anesthesiolog y 6 I: 276-310. 1984 Durant PAC. Yaksh TL. Epidural injections ofbupi vacaine. morphine . fentanyl. lofentanil and DADL in chronicall y implanted rats: a pharmacologic and pathologic study. Anesthesiology 64: 43-53. 1986 Gustafsson LL. Spinal opiate analgesia: clinical and experimental aspects . Thesis . Karolinsk a Institute. Stockholm . 1985 Gustafsson LL. Post C. The degree of analgesia correlates to spinal morphine concentrations after intrathecal administration in rats. Acta Pharmacologica et Toxicologica 58: 243-248. 1986 Gustafsson LL. Post C, Edvardsen B. Ramsay CH. Distribution of morph ine and meperidine after intrathecal adm inistration in rat and mouse . Anesthesiology 63: 483-489. 1985
Spinal Opioid Analgesia
Gustafsson LL, Schildt B, Jacobsen K. Adverse effects of extradural and intrathecal opiates: report of a nationwide survey in Sweden. British Journal of Anaesthesia 54: 479-986, 1982 Herz A, Teschemachen HS. Activities and sites of anti nociceptive action of morphine-like analgesics and kinetics of distribution following intravenous, intracerebral and intraventricular application . Advances in Drug Research 6: 79-119, 1971 Lanz E, Theiss D, Riess W, Sommer E. Epidural morphine for postoperative analgesia: a double-blind study. Anesthesia and Analgesia 61: 236-240, 1982 Le Bars D, Guilbaud G, Jurna I, Besson JM. Differential effects of morphine on responses of dorsal horn lamina V-type cells elicited by A and C fiber stimulation in the spinal cat. Brain Research 115: 518-524, 1976 Malone BT, Beye R, Walker J. Management of pain in the terminally ill by administration of epidural narcotics. Cancer 55: 438-440, 1985 Millan MJ. Multiple opioid systems and pain. Pain 27: 303-347, 1986 Nordberg G. Epidural versus intrathecal route of opioid administration. International AnesthesiologyClinics 24: 93-111, 1986 Nordberg G, Hedner T, Mellstrand T, Dahlstrom B. Pharmacokinetic aspects of epidural morphine analgesia. Anesthesiology 58: 545-551, 1983 Nordberg G, Hedner T, Mellstrand T, Dahlstrom B. Pharmacokinetic aspects of intrathecal morphine analgesia. Anesthesiology 60: 448-454, 1984 Przewlocki R, Stala L, Greczek M, Shearman GT, przewlocka B, et al. Analgesic effects of rnu-, delta- and kappa-opiate agonists and , in particular, dynorphin at the spinal level. Life Sciences 33 (Suppl. I): 649-652, 1983 Rawal N, Arner S, Gustafsson LL, Allvin R. Present state of extradural and intradural opioid analgesia in Sweden: a nationwide follow-up survey. British Journal of Anaesthesia 59: 791799, 1987 Rawal N, Mollefors K, Axelsson K, LindgArdhG, Widman B. An experimental study of urodynamic effects of epidural morphine and of nalaxone reversal. Anesthesia and Analgesia 62: 641-647, 1983 Rawal N, Sjostrand UH, Dahlstrom B, Nydahl PA, Ostelius J. Epidural morphine for postoperative pain relief: a comparative study with intramuscular narcotic and intercostal nerve block. Anesthesia and Analgesia 61: 93-98, 1982 Rawal N, Wattwil M. Respiratory depression after epidural morphine - an experimental and clinical study. Anesthesia and Analgesia 63: 8-14. 1984 Sawe J. Oral morphine and methadone in the treatment of cancer pain: clinical pharmacokinetic studies . Thesis, Karolinska Institute, Stockholm , 1984 Schmauss C, Yaksh TL. In vivo studies on spinal opiate receptor
603
systems mediating antinociception. II. Pharmacological profiles suggestinga differential association of mu, delta and kappa receptors with visceral chemical and cutaneous thermal stimuli in the rat. Journal of Pharmacology and Experimental Therapuetics 228: 1-12. 1984 Sjostrom S, Hartvig P, Tamsen A. Patient-controlled analgesia with epidural morphine and pethid ine. British Journal of Anaesthesia 60: 358-366, 1988 Stenseth R, Sellevold 0 , Breivik H. Epidural morphine for postoperative pain: experience with 1085 patients. Acta Anaesthesiologica Scandinavica 29: 148-156, 1985 Tamsen A, Gordh T. Epidural clonidine produces analgesia. Lancet II: 231-232, 1985 Terenius L. Families of opioid peptides and classes of opioid receptors. In Fields et al. (Eds). Advances in pain research and therapy 9, pp. 461-477, Raven Press, New York, 1985 Yaksh TL. Spinal opiate analgesia characteristics and principles of action . Pain t I: 293-346, 198I Yaksh TL, Harty GS, Onofrio BM. High doses of spinal morphine produce a nonopiate receptor-mediated hyperesthesia: clinical and theoretic implications. Anesthesiology 64: 590-597, 1986a Yaksh TL, Noueihed R, Durant PAC. Studies of the pharmacology and pathology of intrathecally administered 4-anilinopiperidine analogues and morphine in the rat and cat. Anesthesiology 64: 54-66, 1986b Yaksh TL, Reddy SVR. Studies in the primate on the analgesic effects associated with intrathecal opiates, alfa-adrenergic agonists and baclofen. Anesthesiology 54: 451-467, 1981 Yaksh TL, Rudy TA. Analgesia mediated by a direct spinal action of narcotics. Science 192: 1357-1358, 1976 Yaksh TL, Rudy TA. Studies on the direct spinal action ofnarcotics in the production of analgesia in the rat. Journal of Pharmacology and Experimental Therapuetics 202: 411-428, 1977 Yeung JC, Rudy TA. Multiplicative interactions between narcotic agonisms expressed at spinal and supraspinal sites of anti nociceptive action as revealed by concurrent intrathecal and intracerebroventricular injections of morph ine. Journal of Pharmacology and Experimental Therapeutics 215: 633-642, 1980 Weddel SJ, Ritter RR. Serum levels following epidural adm inistration of morphine and correlation with relief of postsurgical pain. Anesthesiology 54: 210-214, 198I
Authors' address: Dr Lars L. Gus/alison, Department of Clinical Pharmacology, Huddinge University Hospital, S-141 86 Huddinge (Sweden).