RESUMEN
OBJECTIVE: To compare the antinociceptive effects of magnesium sulphate (MgSO(4)) when administered epidurally alone and in combination with morphine. STUDY DESIGN: Experimental, randomized, 'blinded', crossover study. ANIMALS: Six healthy adult Beagle dogs. METHODS: Evaluated treatments were MgSO(4) (2.5 mg kg(-1)) alone (Mg), morphine (0.1 mg kg(-1)) alone (Mo), MgSO(4) in combination with morphine (Mm), and sterile water (0.115 mL kg(-1) ; Co) that were injected in the lumbosacral epidural space using an epidural catheter. Antinociception was measured using the von Frey mechanical threshold device applied to the carpal pads, both sides of the thorax and metatarsi. Measurements were obtained at time points: before treatment (baseline) and 0.5, 1, 2, 4, 6, 12, 18 and 24 hours after the epidural injection. Sedation, behaviour score and presence of motor deficits were assessed. Data were analyzed using a linear mixed model and Bonferroni adjustments, with significance set at p < 0.05. RESULTS: There were significant effects of treatment and time in all regions. Overall threshold values in grammes force [median (interquartile range)] when stimulation regions were combined were significantly higher in Mg [164 (135-200)], Mo [156 (129-195)] and Mm [158 (131-192)] compared to Co [145 (120-179)]. Thresholds were significantly higher compared to Co in Mg, Mo and Mm at the thorax and metatarsi, but only in Mg and Mo at the carpal pads. No motor deficits were observed at any time point. Thresholds (combined regions) were increased from baseline at one or more time points with all treatments, including control. CONCLUSION AND CLINICAL RELEVANCE: Epidural MgSO(4) produced an antinociceptive effect characterised by an increase in the mechanical thresholds of similar magnitude to that produced by epidural morphine, compared with the control group, without causing any motor deficits. No potentiation of morphine antinociception was observed. The onset and offset times of antinociception could not be clearly established. To what extent these results can be extrapolated to clinical cases requires further investigation.