Key words: Anesthetics, local: lidocaine; systemic. Chronic pain: allodynia.
In science, medicine, and politics, simple explanations for complex problems are desired. Basic reasons for this tendency toward parsimony in human behavior are that the simple forms are, respectively, easier to understand and to test, easier to practice, and easier to justify to the electorate. Counter to this notion is the stimulation we receive from the unexpected, the complex, and the unpredictable. There are many more possibilities and choices, accompanied by ample confusion. Who among us has not been disconcerted from a reality intrusion that challenged our assumptions? A very unfortunate few, I hope, for much education, wisdom, and even delight can ultimately result from the initial surprise of the unanticipated.
In this issue of ANESTHESIOLOGY, we are called to reconsider the fundamental pharmacology of local anesthetics. Chaplan et al. report alleviation of hyperalgesia of a week's duration that follows an hour's infusion of systemic lidocaine, thereby relieving a previously induced peripheral neuropathy, the behavioral symptoms of which resemble allodynia in humans. Analgesia, assessed by the quantitatively determined restoration to control values of mechanical threshold for limb withdrawal, was well correlated with the plasma lidocaine concentrations reached during the brief infusion. Neither intrathecal nor peripheral blockade of nerve at these times resulted in this prolonged analgesia.
These findings are consequential in three respects. First, they demonstrate that the transient presence of the drug disrupts the behavior for a much longer time, a phenomenon long known from the clinical literature. [2-5]This may arise from the transient suppression of "reverberating" neural circuits, a form of forgetting, from the activation of some prolonged, endogenous analgesic source or from the presence of a quasistable metabolite of lidocaine in a compartment where it is sequestered yet available for biologic activity.
Second, they allow that the central and not the peripheral nervous system may contain the site(s) of action. Whereas it is very unlikely that the doses of lidocaine injected intrathecally or around the sciatic nerve yielded measurable systemic concentrations, the systemic infusions can, in principle, result in actions in the periphery as well as centrally. Contrary to this possibility is the finding that systemic lidocaine is a more effective analgesic for peripheral neuropathic pain than for pain arising in the central nervous system. [7,8]Systemic local anesthetics thus provide a diagnostic as well as a therapeutic tool. [7,9]
The third noteworthy aspect questions the mechanism for lidocaine's analgesic effect. We often and automatically associate local anesthetics with the blockade of propagating nerve impulses, mediated through the direct inhibition of voltage-gated sodium channels. A spectrum of agents that are effective at Sodium sup + channels and intended for different therapeutic ends are known to provide relief from neuropathic pain when given systemically. Blockade of high-frequency bursts of impulses may be required for this effect, however. Blockade of single rat peripheral fibers in vivo requires more than 0.1-mM concentrations of lidocaine, whereas the analgesic plasma lidocaine EC50s measured by Chaplan et al. are about 0.003 mM (0.75 micro gram/ml). Such low concentrations will bind to fewer than 2% of the Sodium channels in a resting axon and probably to less than 10% of these channels in an axon firing impulses at high physiologic frequency. Others have reported analgesic lidocaine plasma concentrations in humans of 20-40 micro Meter, a range where glutamate- and tachykinin-mediated spinal cord responses to C-fiber evoked activity are selectively depressed while compound action potentials are unaffected. 
There are alternative explanations for the observed analgesia. One is that the primary action is through the inhibition of sodium channels but that the site is at regions of very low conduction safety, such as the distal generator for spike or a central, presynaptic arborization of a primary afferent fiber where many branch points divide the local circuit current to near marginal levels. 
It is important to realize that, even if the local anesthetic actions occur on Sodium sup + channels, the primary lesion may not produce changes in Sodium sup + channels. Any condition that leads to partial steady-state depolarization or to high-frequency impulse firing renders a membrane more susceptible to local anesthetic blockade. An example of this occurs in diverse forms of myotonia, where prolonged muscle contraction arises from repetitive firing of sarcolemmal action potentials due to molecular defects in either chloride channels or Sodium sup + channels. Despite these different molecular origins, myotonic symptoms are similarly treatable with systemic local anesthetics.
A nontraditional explanation is that these analgesic effects arise from targets other than sodium channels. Historically, many such targets, both membrane channels and enzymatic activities, have been identified, but the effective concentrations of local anesthetics often exceed even those for sodium channel inhibition. Recent reports, however, reveal that both excitatory, glutamate-gated synaptic potentials in spinal cord and peptide hormone-mediated inhibition of ionic potassium current in cultured central neurons are depressed by these "analgesic" concentrations of lidocaine.
A third explanation is essentially pharmacokinetic. The anesthetic may reach peripheral neuronal sites accessible from the circulation but not from direct extraneural application, or a slowly degraded metabolite of the anesthetic, produced locally or systemically, may be the active agent.
How should we respond to these observations? There are new possibilities in science and in practice. Scientifically, let us consider the possibility that the actions of systemic as well as epidural and intrathecal local anesthetics likely involve targets other than sodium channels. This will expand our perspective of local anesthetic mechanisms and perhaps lead to new investigations of local anesthetics on the neuraxis. Practically, let us reconsider the use of low, nontoxic concentrations of systemic lidocaine for prolonged periods to accomplish peri- and postoperative pain relief. Perhaps this application will contribute to an effective preemptive analgesia. Clinical studies can be conducted in a safe, controlled way to explore this possibility.
Gary Strichartz, Ph.D., Professor of Anaesthesia/Pharmacology, Vice Chairman for Research, Anesthesia Research Laboratories, Brigham and Women's Hospital, 75 Francis Street, Boston, Massachusetts 02115.