Potent analgesic effects of anticonvulsants on peripheral thermal nociception in rats

Abstract

1. Anticonvulsant agents are commonly used to treat neuropathic pain conditions because of their effects on voltage- and ligand-gated channels in central pain pathways. However, their interaction with ion channels in peripheral pain pathways is poorly understood. Therefore, we studied the potential analgesic effects of commonly used anticonvulsant agents in peripheral nociception. 2. We injected anticonvulsants intradermally into peripheral receptive fields of sensory neurons in the hindpaws of adult rats, and studied pain perception using the model of acute thermal nociception. Commonly used anticonvulsants such as voltage-gated Na+ channel blockers, phenytoin and carbamazepine, and voltage-gated Ca2+ channel blockers, gabapentin and ethosuximide, induced dose-dependent analgesia in the injected paw, with ED50 values of 0.30, 0.32 and 8, 410 microg per 100 microl, respectively. 3. Thermal nociceptive responses were not affected in the contralateral, noninjected paws, indicating a lack of systemic effects with doses of anticonvulsants that elicited local analgesia. 4. Hill slope coefficients for the tested anticonvulsants indicate that the dose-response curve was less steep for gabapentin than for phenytoin, carbamazepine and ethosuximide. 5. Our data strongly suggest that cellular targets like voltage-gated Na+ and Ca2+ channels, similar to those that mediate the effects of anticonvulsant agents in the CNS, may exist in the peripheral nerve endings of rat sensory neurons. Thus, peripherally applied anticonvulsants that block voltage-gated Na+ and Ca2+ channels may be useful analgesics.

Figure 1

Phenytoin and carbamazepine induce analgesia in thermal PWL testing. Upper panel: Injections of 1% DMSO in saline (open symbols), which was used as a vehicle for experiments with phenytoin and carbamazepine, did not significantly change PWLs in the injected side in comparison to the noninjected side (filled symbols). Middle panel: Phenytoin induces a dose-dependent increase in thermal PWL. Phenytoin (0.3, 1.0 and 3.0 μg) significantly increased PWLs (*, F(1, 22)=13.51, P=0.001; F(1, 22)=119.83, P=0.000; and F(1, 20)=24.53, P=0.000, respectively) at 10 min and (*, F(1, 22)=9.97, P=0.005; F(1, 22)=24.327, P=0.000; and F(1, 20)=28.4, P=0.004, respectively) at 20 min postinjection when compared to the noninjected paw (control, open circles). PWLs returned to control values by 60 min following injection; control (noninjected) sides are grouped together. Lower panel: Carbamazepine induces a dose-dependent increase in thermal PWL. Carbamazepine (0.2, 0.7 and 2.0 μg) significantly increased PWLs (*, F(1, 16)=19.02, P=0.000; F(1, 16)=65.08, P=0.000; and F(1, 16)=25.16, P=0.000, respectively) at 10 min and the two highest doses (0.7 and 2.0 μg) significantly increased PWLs (*, F(1, 16)=12.80, P=0.003 and F(1, 16)=14.30, P=0.002, respectively) at 20 min postinjection when compared to the noninjected paw (control, open circles). Again, PWLs returned to control values by 60 min following injection.

Figure 2

Gabapentin and ethosuxamide induce analgesia in thermal PWL testing. Upper panel: Injections of saline alone (open symbols) did not significantly change PWLs when compared to noninjected paws (filled symbols). Middle panel: Gabapentin induces a dose-dependent increase in thermal PWL. Gabapentin (50 and 170 μg) significantly increased PWLs (*, F(1, 16)=5.56, P=0.031 and F(1, 16)=138.14, P=0.000, respectively) at 10 min and the highest dose (170 μg) significantly increased PWLs (*, F(1, 16)=38.64, P=0.000) at 20 min postinjection when compared to the noninjected paw (control, open circles). PWLs returned to control values by 60 min following injection. Lower panel: Ethosuxamide induces a dose-dependent increase in thermal PWL. Ethosuxamide (420 and 1400 μg) significantly increased PWLs (*, F(1, 16)=22.74, P=0.000 and F(1, 22)=175.81, respectively) at 10 min and the highest dose (1400 μg) significantly increased PWLs (*, F(1, 16)=8.54, P=0.010) 20 min postinjection when compared to the noninjected paw (control, open circles). PWLs returned to control values by 60 min following injection.

Figure 3

Dose-dependent local thermal analgesia following local injections of anticonvulsants. Average dose–response curves for local analgesia indicated by maximal increase in thermal PWLs on injected side vs noninjected side 10 min after injection (y-axis) and dose of anticonvulsant expressed in μg 100 μl⁻¹ (x-axis). All points are normalized to the maximal effect seen with a particular agent from data presented in Figure 1 and Figure 2 and are averages of at least nine experiments. Solid lines are best fits of the Hill equation (see Methods) and downward vertical lines indicate s.e.m. Fits were constrained to 100% block with ED₅₀ and h values of 0.30±0.02 μg (h=2.8±0.9), 0.32±0.04 μg (h=2.1±0.4), 8±1 μg (h=0.9±0.1) and 410±37 μg (h=2.5±0.7) for phenytoin, carbamazepine, gabapentin and ethosuximide, respectively.