Interspecies scaling: A comparative assessment

Interspecies (Allometric) Scaling is the study of the change in proportion of various parts of an organism as a consequence of growth. For over two decades, allometry has been used to predict human pharmacokinetic parameters from preclinical animal data. Presently, five distinct prediction methods are available for pharmaceutical scientists to perform allometric scaling. Simple allometry acts as the blueprint for three of the five methods and of these most widely accepted is the Rules of Exponents. Other practices outlined by leading scientists and the FDA provide direction for using allometry with fixed exponents. The Two Species Method created by provides ways to predict human clearance using rats and dogs or monkeys, while One Species and the Liver Blood Flow Method use a single species to generate its predictions. The One Species Method uses optimization of its coefficients to generate predications while the Liver Blood Flow Method uses the animal clearance and the ratio of hepatic flow of humans to mammals.;The goal of this dissertation research was to conduct a comprehensive retrospective comparison of the various Interspecies Scaling methods used to predict drug clearance in humans. The research addresses the following specific aims: (1) To evaluate the extent to which each Interspecies Scaling method successfully predicts human clearance, (2) To determine if the Rule of Exponents is a better predictor of human clearance in comparison to other allometric methods, (3) To assess the effect of reducing the number of species on the predictive ability of these approaches, and (4) To determine whether the ratio of predicted human clearance to observed animal clearance (Predicted to Observed Ratio) provides enough evidence to distinguish between predictive success and failure. Although this is a retroactive comparison, a fourth aim (4) was to determine if predicative success can be established before conducting first-in-humans trails based on the ratio between the predicted human clearance and species clearance.;Using basic spreadsheet software, 113 compounds were analyzed by fourteen Interspecies Scaling approaches. An acceptable range for success was defined as any prediction within +/-75% of observed human clearance. The success of each method was evaluated by percent error (between the observed and predicted human), ratio of predicted to observed values, and the frequency that human clearance was predicted successfully. Each compound was labeled excellent, good, fair, poor or failure. In a subsequent study, a "Best Methods" comparison was devised to compare the methods based on the number of species used, overall predictive strength, overall percent error and whether the method was a better predictor than the Rule of Exponents. The findings from the Best Method comparison of Data Set 1 were used to evaluate Data Set 2.;Interspecies Scaling approaches were validated by the ∼80% predictive strength success rate. Despite being the standard scaling technique, the Rules of Exponents was proven not to be the optimal method. Reducing the number of species had an ambiguous effect on predicative success. Typically selecting the monkey for single species approaches increased method predictability, while selecting rat or dog decrease predictive success, as compared to three species methods. Predicative success could not be established based on the ratio between the predicted human clearance and species clearance. Using a combination of three, two and single species methods provides the user with a panoramic clearance profile.