In animal models of neuropathic pain, transection or constrictive injury to peripheral nerves produces ectopic discharges originating at both injury sites and related dorsal root ganglia (DRG). In addition, hyperexcitability is observed in associated dorsal horn (DH) neurons of the spinal cord. As ectopic discharges are inhibited by agents that block voltage-sensitive Na+ channels, it has been postulated that accumulation of Na+ channels in the membrane at nerve injury sites may contribute to the development of ectopic nerve activity (ENA). The goal of the present study was to compare the sensitivity of ENA to lidocaine and QX-314, a positively charged lidocaine derivative, which is frequently assumed to be membrane impermeant. Experiments were performed on adult male Sprague-Dawley rats in which the common sciatic nerve had been transected 4-10 days earlier. Extracellular microelectrode recordings were made from DRG and DH neurons, and neuronal activity was measured in fine bundles of microfilaments teased from sciatic nerves in anesthetized and paralyzed rats. Comparative effects on heart rate (HR) and mean blood pressure (MBP) were also studied. To confirm that externally applied QX-314 is able to inhibit high frequency activity in sensory nerves, QX-314 was superfused over isolated rat vagus nerves during stimulation of compound action potentials in C-fibers (C-spikes). As expected, intravenously administered lidocaine inhibited ENA at all three sites. Lidocaine ED50 values (expressed as mg/kg, with 95% confidence limits) were: 10.2 (7.8-13.3), 1.4 (0.8-2.4) and 0.9 (0.4-2.0) for neuromas, DRG and DH neurons, respectively. QX-314 also induced dose-dependent inhibition of ENA at neuromas and DRG, but produced only a small inhibition of DH neuron ENA. QX-314 had the following ED50 values (mg/kg) for neuromas, DRG and DH neurons, respectively: 2.3 (2.0-2.8), 6.9 (4.7-26.5) and 85.7. QX-314-mediated inhibition of DRG ENA had a slow onset and was long-lasting, relative to lidocaine. Lidocaine or QX-314 also significantly reduced HR and MBP in the same dose range as that which reduced ENA in DRG or neuromas. In isolated rat vagus nerve recordings, QX-314 induced marked use-dependent inhibition of C-spike amplitude, with IC50 values (microM) of 9000 (4600-18,000) and 350 (290-420) for low- (0.03 Hz) and high-frequency (30 Hz) C-spikes, respectively. These data support the hypothesis that Na+ channel accumulation contributes to the generation of ectopic discharges in neuromas and DRG, and suggests that intravenous QX-314 can acutely block Na+ channels at these sites.