Intrathecal midazolam and fentanyl in the rat: evidence for different spinal antinociceptive effects.

The effects of intrathecal midazolam and fentanyl on electrical current threshold for pain were measured using stimulating electrodes in the neck and tail of rats with chronically implanted lumbar subarachnoid catheters. This involved the measurement of the minimum current (50 Hz 2 ms pulses 0-5 mA), which made the rat squeak when applied alternately to electrodes at each skin site. The responses measured in milliamperes were expressed as a number of times control readings. Equieffective doses of both midazolam and fentanyl produced a significant increase in electrical threshold for pain in the tail (mean +/- SEM 3.14 +/- 0.51 and 2.89 +/- 0.35: P less than 0.05; Wilcoxon sum rank test) in the absence of any change in the neck (mean +/- SEM 1.28 +/- 0.13 and 0.96 +/- 0.12, NS), thus demonstrating a spinal effect. Fentanyl caused a significant simultaneous increase in tail flick latency (mean +/- SEM 67.8 +/- 20.1%, P less than 0.05), but midazolam did not (mean +/- SEM 4.22 +/- 2.76%, NS). Intraperitoneal injections of naloxone (0.25 mg/kg) blocked the response to fentanyl in both tests and did not affect the response to midazolam. Intraperitoneal flumazenil (5 mg/kg) blocked the midazolam antinociceptive effect but did not affect the response to fentanyl in either test. Tail withdrawal in response to non-noxious stimulation was preserved in all animals with spinal analgesia, indicating that myelinated afferent and efferent pathways were still functioning. Righting reflex, coordination, motor power, and alertness were also preserved in the presence of both drugs. Both drugs caused spinally mediated antinociceptive effects that were qualitatively different.