Duration of opioid antagonism by nalmefene and naloxone in the dog. A nonparametric pharmacodynamic comparison based on generalized cross-validated spline estimation.

The opioid antagonist nalmefene was compared in its pharmacodynamic properties to the structurally similar antagonist naloxone in a 2 x 2 cross-over study with 8 dogs. Opioid-induced respiratory depression was produced for ca. 7 hours with a constant rate intravenous infusion of 30 micrograms/kg/hr fentanyl and quantified using noninvasive transcutaneous pCO2 recordings. Upon reaching a pseudo-steady state of respiratory depression at 2 hours post fentanyl infusion initiation, the animals then received either nalmefene (12 micrograms/kg/hr) or naloxone (48 micrograms/kg/hr) for 30 minutes. The pharmacodynamic pCO2 responses produced by the combined agonist/antagonist regimen were fitted with a cubic spline function using a generalized cross-validation technique. Various quantities that describe the onset, duration and relative potency of each antagonist were determined directly from the estimated response curves in a model-independent, nonparametric way. The 2 antagonists were compared in terms of these quantities using a statistical model that considers carry-over effects typically arising from a possible development of tolerance. The results indicate that nalmefene: 1. is approximately 4-fold more potent than naloxone, 2. has an onset of reversal as rapid as naloxone, and 3. has a significantly longer (2-fold) pharmacodynamic duration of action than does naloxone. The mean time required for the agonist to regain 30% or 50% of its effect present at the start of the antagonist infusion was 66 and 112 minutes and 37 and 55 minutes for nalmefene and naloxone, respectively. Early, effective pharmacodynamic screening of new drug compounds is a valuable way of accelerating the drug discovery process and reducing escalating drug development costs. This study exemplifies a novel, endpoint oriented pharmacodynamic comparison procedure that can be done expeditiously before starting the time consuming development and validation of a drug level assay, and before engaging in considerably more involved integrated PK/PD studies.