Objective and design: Relevance of the preclinical pharmacodynamic, toxicity and pharmacokinetic parameters predicting the clinical potency of nonsteroidal antiinflammatory drugs (NSAIDs) was evaluated.
Material: Data for oral potencies of 24 NSAIDs in rats were collected from the literature and from New Drug Applications with respect to the following parameters: antiinflammatory, analgesic, antipyretic, acute ulcerogenic activities, acute toxicity, in vitro inhibition of prostaglandin synthesis, acid dissociation constant (pKa), octanol-water partition coefficient and elimination half-life.
Treatment: Data for most of the in vivo parameters in rats were collected following single dose administration with the exception of adjuvant arthritis. Single and daily clinical doses were considered. All of these NSAIDs have been approved for marketing although not all have been sold in the USA.
Methods: The preclinical data were compared to human dose (unit or daily doses) using single and multiple stepwise regression analyses.
Results: Analyses suggest that NSAIDs are effective in all models of preclinical tests for fever, pain and inflammation, however, carrageenin-induced rat paw edema model is clearly the best predictor of human dose. Rank order of preclinical models for predicting human dose is carrageenin > yeast induced fever > pressure induced pain = adjuvant arthritis in rats. The analysis suggested that the pain and adjuvant arthritis models in rats may also involve a prostaglandin independent mechanism. Of the two physicochemical factors tested, pKa contributed best to the carrageenin model towards predicting the clinical potency of NSAIDs. Mathematical relationships between human dose, carrageenin ED50 and pKa were established that may assist in the future clinical development of NSAIDs.
Conclusions: Carrageenin-induced paw edema model in rats is the most robust predictor of the clinical potency of NSAIDs. Acid dissociation constant (pKa) appears to be a secondary contributor to the potency of NSAIDs. The relevance of the data analyses for developing cyclooxygenase-2 (COX-2) selective NSAIDs is discussed.