Antinociceptive effects of epidural magnesium sulphate alone or in combination with morphine in dogs

The analgesic properties of magnesium meditated by its physiological antagonistic action on the NMDA-receptor is of great interest in human and veterinary medicine. The primary objectives of this study were to investigate whether the lumbosacral epidural injection of magnesium could produce an antinociceptive effect and to determine whether there was possible potentiation of the antinociceptive effect between magnesium and morphine when administered epidurally in combination in dogs. A secondary goal was to study the onset and duration of the antinociceptive effect of epidural magnesium alone or in combination with morphine in dogs. Furthermore, the possible motor deficits induced by epidural magnesium were investigated. Six healthy, adult, neutered research Beagle dogs (3 male and 3 female) were used in a randomized blinded crossover study with a one-week wash-out period between treatments. Treatments consisted of an epidural injection of: 0.115 mL kg-1 of sterile water (treatment Co); 0.1 mg kg-1 of morphine (treatment Mo); 2.5 mg kg-1 of MgSO4 50% (treatment Mg); and 2.5 mg kg-1 of MgSO4 together with 0.1 mg kg-1 of morphine (treatment Mm). Sterile water was added to treatments Mo, Mg and Mm to receive a total volume of 0.115 mL kg-1. Dogs were anaesthetized with propofol and isoflurane to place a lumbosacral epidural catheter for the administration of the treatments. Antinociceptive effects were evaluated at different time points for 24 hours post-injection using von Frey mechanical thresholds. Three threshold measurements on both sides were obtained at the carpal pads, thorax and metacarpi at each time point and then averaged for statistical analysis. Maximum applied force eliciting a nociceptive response was recorded and compared between treatments. Within each treatment, measurements obtained at different time points were compared with baseline values. Tail tone, level of ataxia, level of sedation and behaviour were scored at each time point. Data were analysed using a linear mixed model with significance set as p