The trusted source for
healthcare information and
Source: Samad TA, et al. Interleukin-1beta-mediated induction of Cox-2 in the CNS contributes to inflammatory pain hypersensitivity. Nature. 2001;410:471-475.
Cyclooxygenase (Cox)-2 inhibitors such as celecoxib (Celebrex) and rofecoxib (Vioxx) have become popular drugs not only for the treatment of arthritis, but for a wide range of pain syndromes. The data of Samad and colleagues indicate that this clinical practice may have a basis in science. In addition to inhibition of the peripheral production of prostaglandins, suppression of Cox-2 within the spinal cord and brain may explain the potential efficacy of these agents.
Traditional NSAIDS act nonspecifically on both isoforms of Cox: Cox-1, which is expressed constitutively in all tissues including the gastric mucosa and kidney; and Cox-2, which is specifically induced at the site of inflammation. Selectivity for Cox-2, therefore, maximizes the therapeutic benefit of these drugs while avoiding side effects such as gastrointestinal bleeding. Cox enzyme activation promotes the conversion of arachidonic acid to prostaglandins. In the peripheral nerve terminals, prostaglandins then activate protein kinases, which lead to excitation of sodium channels and reduction of pain thresholds. As Samad et al demonstrate, Cox-2 and prostaglandin activity also contribute to hypersensitivity of pain pathways in the spinal cord and brain.
Using a well-accepted rat model of hindpaw inflammation induced by injection of complete Freund’s adjuvant (CFA), Cox-2 mRNA was measured in the lumbar spinal cord at the L4-5 segments. There was a 16-fold increase in mRNA expression at 6 hours, ipsilateral to the stimulus. Bilateral Cox-2 activation was observed extending to 24 hours after the onset of pain. Increases in Cox-2 mRNA expression were also observed in the cervical spinal cord, pons, ventral midbrain, and hypothalamus; a large 12-fold sustained increase occurred in the thalamus. This central Cox-2 induction was accompanied by greater than 80-fold increase in prostaglandin E2 (PGE2) concentration in the CSF.
Samad et al test 2 possible hypotheses for the mechanism of Cox-2 induction in the CNS: sensory inflow from nerve fibers innervating the inflamed hindlimb or activation by circulating pro-inflammatory cytokines. The first of these possibilities was tested using an injection of bupivacaine to induce complete sensory and motor blockade of the sciatic nerve. Samad et al demonstrate that Cox-2 induction and PGE2 production are reduced but not eliminated. Interestingly, C-fiber activation using an electrical stimulus induced Cox but at much lower levels than did inflammation. As Samad et al contend, these data make a transsynaptic mechanism much less likely and point to an inflammation-related, likely humoral, mechanism of Cox expression. A prime candidate for this is the cytokine IL-1•, which is upregulated more than 10,000-fold after the hindpaw is inflamed.
Receptors for IL-1• are highly expressed in the spinal cord, particularly Rexed laminae.1-3 Interestingly, exogenous administration of IL-1• by an intrathecal route produced 20- to 30-fold more Cox activation than did an equipotent intravenous dosage. Endogenous IL-1• is likely produced not only at the site of peripheral injury but also by glial or neuronal cells within the spinal cord itself. Given that IL-1• levels are not found to be elevated in the serum following hindlimb inflammation, CNS induction of IL-1• may require an additional second messenger, such as IL-6.
In a behavioral model, using calibrated Von Frey filaments to test sensitivity to mechanical irritation, Samad et al show that in the setting of Cox-2 activation, rats were much more sensitive to painful stimuli. These hyperalgesic responses were blocked by intrathecal injection of the Cox-2 antagonist NS 398 or injection of YVAD, an inhibitor of IL-1• production. Intravenous injection did not produce these effects. Animals not subjected to hindpaw inflammation, who did not have upregulation of Cox-2, showed no effect from these inhibitors.
As Samad et al observe, the constitutional symptoms associated with inflammation and infection—such as fever, lethargy, malaise, and anorexia—are also related to increases in IL-1• and Cox-2. Inhibition of IL-1• production or administration of anti-inflammatory cytokines attenuate these effects.
As these data indicate, inhibiton of Cox-2 activity, either with selective drugs such as celecoxib or nonspecific NSAIDs, may significantly modify central pain pathways. But are these effects clinically significant? Since only a fraction of these drugs cross the blood brain barrier, it is unlikely that orally administered drug has a significant effect on these CNS processes. Cox-2 inhibitors should, thus, be designed to penetrate into the brain and spinal cord and, therefore, target central receptors. Other strategies might include downregulation of Cox gene expression by targeting IL-1• production or blocking IL-1• receptors. Because the current formulation of drugs such as celecoxib offers no particular benefit in this regard, agents such as ibuprofen, which are significantly less expensive and also affect Cox-2, should be considered first-line therapy in patients who tolerate them.
We should caution that these data should not be interpreted as evidence that Cox-2 plays a major role in pain that is neuropathic in nature (eg, postherpetic neuralgia or reflex sympathetic dystrophy) or in nociceptive pain that is purely mechanical without a significant inflammatory component (allodynia). As Samad et al’s data clearly indicate, animals not subjected to the inflammatory injury did not show benefit from Cox inhibition and activation of C-fibers using electricity did not mimic CFA-associated Cox upregulation.
Thus, these data are not an indication that Cox-2 inhibitors have activity beyond that of an anti-inflammatory. Rather, they indicate that inflammation is very much a CNS process. The cascade of cytokine induction, Cox activity, and prostaglandin production directly affects the spinal cord and brain. Because there may be important interactions between Cox and other mediators of central pain sensitization such as glutamate and substance P, it is likely that the ideal strategy for pain control will target multiple neurochemical substrates. —Alan Z. Segal