| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesiology, University of California, San Diego, California
Address correspondence and reprint requests to Tony L. Yaksh, PhD, Department of Anesthesiology, University of California, 9500 Gilman Dr., La Jolla, CA 92093-0818.
In their case report, Soliman et al. (1) sought to use baclofen as an analgesic adjuvant in children undergoing elective dorsal rhizotomy for the management of spasticity. At the doses used, there was an amelioration of pain, and this was accompanied in two patients by transient (2 h) paralysis and, in one case, sensory loss.
The rationale behind the use of baclofen, a 
-aminobutyric acid- (GABA) B agonist, is based on a prevalent preclinical literature that has shown that the spinal GABA-B receptor can regulate the response to acute high-intensity input (2–4), the facilitated pain state that is generated by tissue injury (5) and the component of the hyperalgesia and allodynia that arises from nerve or spinal injury (6–8). Importantly, these antinociceptive effects have also been observed in primate behavioral models (9). The mechanism of these actions is not fully understood, but the intrathecal agonist and antagonist pharmacology indicates the role of spinal GABA-B receptors.
In the spinal cord, GABA-B binding is present throughout the spinal gray matter with a maximal concentration in Lamina II of the dorsal horn. Dorsal rhizotomy reduces dorsal, but not ventral, horn GABA-B binding (10). Binding studies have shown high levels of such binding in the substantia gelatinosa (see above references). GABA-B binding has been demonstrated in both the dorsal and the ventral gray matter of the spinal cord. Capsaicin treatment results in a reduction in GABA-B binding (11). These observations suggest that a proportion of the GABA-B binding sites are on certain populations of small primary afferents and that GABA-B agonists may regulate the excitability of the primary afferent terminal and reduce the release of transmitter from those terminals. Such a reduction in the evoked release of substance P or calcitonin gene related peptide (12,13) and a reduced internalization of the neurokinin 1 receptors has indeed been identified (14), although others have failed to show such changes in release with moderate baclofen concentrations (15,16). Consistent with this association, spinal baclofen produces a dose-dependent inhibition of the C, but not A, fiber-evoked activity in dorsal wide dynamic range neurons (17). These observations account for effects that may be mediated by small afferent input (e.g., after acute high intensity stimulation and tissue injury). The effects of GABA-B activity at the spinal level are not certain. In the face of transient withdrawal of GABA-A activity, a hyperalgesia and tactile allodynia is clearly evident (18). Blocking the GABA-B receptor has no such allodynic effect. Accordingly, it is not clear whether the loss of GABAergic tone, as reported after nerve injury (19,20), reflects that there is an intrinsic role for GABA-B receptors in regulating the postnerve-injury state.
Of particular interest, GABA-B agonists may regulate the spinal release of excitatory amino acids (21). Such a reduction in glutamate release would be consistent with the preclinical behavioral effects of intrathecal baclofen in the hyperalgesia and allodynia associated with models of nerve and tissue injury (22).
It should be stressed that, although the GABA-B receptor can have potent effects on sensory processing, it significantly affects motor function. Its ability to hyperpolarize motor horn cells and reduce by a presynaptic action the excitatory postsynaptic potentials in motor neurons made by IA sensory afferents (23), as well as the direct hyperpolarization of the second order membrane, likely accounts, in part, for the potent inhibitory effect on motor horn excitability that correlates with the therapeutic effect on spasticity. Even in the first study, which examined the effect of spinal baclofen, it was emphasized that the antinociceptive effects were accompanied at the upper end of the antinociceptive dose effect curve by the observation of hind limb weakness leading to an extreme, but reversible, weakness (2). Thus, there has always been a question of whether the spinal effects on sensory processing could be separated from the changes in motor function.
Given the profound effects observed with GABA-B agonists in the animal models, it has been somewhat puzzling that a comparable profound analgesia has not been routinely reported with baclofen in humans. Baclofen probably ranks as one agent that has been given spinally with great frequency because of the therapeutic advantage that arises from regulating the enhanced motor tone associated with spasticity (24), a clear clinical expression of the preclinical observations made first in 1978. Despite this frequency, it has been difficult to define an analgesic effect in humans. A few reports that have appeared suggest that baclofen may be useful in some neuropathic pain expressions (25–27). As in the rodent studies, it may be argued that i) the pain relief may arise because of the reduction in the pain otherwise generated by the continuously contracted musculature, or ii) the dose required to produce muscle relaxation in spasticity may be substantially smaller than those doses required to alter nociceptive processing and that the clinician is loathe to exceed the therapeutic dose necessary for muscle relaxation to avoid tolerance.
Thus, in the context of the present brief report regarding baclofen and muscle weakness, it would seem that a drug has to do what a drug has to do. The present case series does, however, leave many issues unresolved. It seems clear that the patients did receive significant pain relief, equivalent to that associated with intrathecal fentanyl. This is consistent with at least the growing case literature. The unintended, profound flaccidity is one that provides no insight one way or the other. At the time of surgery, the delivery of a bolus injection of baclofen might be anticipated to effect motor function that is dose-dependent. We do not know the intrathecal distribution of baclofen. It is known that, in adults, the size of the thecal sac can vary significantly between individuals, and that can lead to differing local concentrations of the drug (such as tetracaine) given by that route and different magnitudes of local anesthesia (28). It is possible that, in these spinal cords, such a difference in distributional properties might account for the different and, in some cases, unexpectedly intense physiological effects. Alternatively, it is clear that neurodegenerative disorders may lead to changes in neuronal net function and differential drug sensitivity. In other words, the apparent changes in baclofen sensitivity may reflect a specific characteristic of that spinal cord with which the extreme drug effect was noted. The only manner in which this issue can be approached is by the use of a systematic dose analysis or by an opportunity to titrate the baclofen dosing, which may be achieved with a catheter placed at the time of surgery. This poses an additional opportunity for potential complications.
Other drug modalities that might be theoretically considered include the use of longer lasting opiates such as morphine or methadone, the 
2 agonist clonidine, or an N-methyl-D-aspartic acid (NMDA) antagonist such as ketamine. There are several theoretical arguments for the use of such drugs. The pain states are clearly opiate sensitive (if fentanyl really does produce a spinal action that is sustained when given in this manner). It should be emphasized that opiates can induce a myoclonus, possibly mediated in part by a direct effect on motor horn inhibitory interneurons or through active metabolites (29,30). Methadone is more lipid soluble than morphine, but has a long t1/2ß half-life. Interestingly, the racemic mixture of methadone consists of the active opiate agonist (d-methadone). The L-isomer is not an opiate, but has been shown to have significant NMDA antagonistic properties (31,32). Facilitated pain states secondary to nerve and tissue injury are regulated in part by spinal NMDA receptors (22). Intrathecal d-methadone has antihyperalgesic actions (33). Spinal 2 receptors have a potent effect on small afferent input and are also known to strongly reduce excitatory amino acid release (34). Behaviorally, spinal 2 agonists in animal models (35) and in humans (36) have potent antiallodynic effects. Of additional interest, in animal models, large doses of clonidine will produce muscle weakness and reduce motoneuron excitability in rats with spinal injury (37) and has been used in humans to control spasticity (38,39).
In summary, Soliman et al. (1) probably did, in fact, observe an amelioration of the pain state. The origin of this pain state is clearly complex and it is not clear how the effects of baclofen on pain can be interpreted. Thus, the postoperative pain arises not only from the tissue trauma, but from the acute changes induced by sectioning of the nerve itself. The "side effect" of flaccidity with baclofen is not surprising and little can be made of the sensitivity of these patients to baclofen as only a limited dose range was examined. The lack of information on the kinetics of the spinal space after such a surgery in such children further complicate the issue. Nevertheless, it raises some limited hope that appropriate attention to dosing and to the considerations of other spinal pharmacological approaches might provide an alternative means to address this postneurectomy/postoperative pain state.
References
2 agonists. Peptides 1993;14:371–8.[ISI][Medline]
2-adrenergic agonists as analgesics. In:Aitkenhead AR, Benad G, Brown BR, et al., eds. Baillière’s clinical anaesthesiology. Vol 7. No. 3. London: Baillière Tindall, 1993:597–614.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
