The competitive NMDA antagonist GPI 3000 (GPI) does not improve outcome after cardiac arrest in dog

Introduction: We tested the hypothesis that GPI administered after resuscitation from cardiac arrest would improve 4 day recovery as measured by neuro behavioral assessment and neuropathotogy. Methods: With approval of the Animal Care and Use Committee at JHMI, hatothane/N₂O anesthetized dogs were subjected to 7 minutes of cardiac arrest followed by vest cardiopulmonary resuscitation. Neurobehavkxal outcomes were scored daily and on the fourth day, the animals were euthanized and neuropathotogy was evaluated in hippocampus and neocortex by Hematoxylin and Eosin. Three groups were treated in a randomized blinded protocol with either saline (SAL), GPI-LO (5 mg/kg fotowed by 1 mg/kg/hr for 2 hrs), or GPI-HI (25 mg/kg. followed by 5 mg/kg/hr for 2 hrs). Results: The mortality rate was higher in animals receiving GPI than in saline controls (4/15 deaths SAL, 6/15 in GPI-LO, and 9/18 in GPI-HI). Neurobehavkxal scores were worse in GPI-treatad animals than in saine controls (p<.05, ANOVA) in a dose-dependent manner. Neuropathotogic damage in neocortex was most severe in GPMreated animals (p<.05). with the percent of injured neurons dependent upon the dose: 8.3 ±2.7% SAL, 13.2 ± 6.4% GPI-LO, and 39.4 ±10.1% for animals treated with GPI-HI. CA1 neuronal damage was severe in all groups. Conclusions: Contrary to results in global and focal cerebral ischemia, NMDA receptor antagonism worsens outcomes after cardiac arrest and resuscitation in dog.