There may be circumstances in which dose extrapolation among species is linear, that is, the most appropriate interspecies comparison is based on scaling directly to body weight, i.e., NOAEL (mg/kg) = HED (mg/kg). For example, if the NOAEL in each animal species is similar on an mg/kg basis, then extrapolation to humans based on mg/kg may be acceptable. However, since such a similarity could result from factors unrelated to drug sensitivity (e.g., differences in absolute oral bioavailability), if data from only two animal species are available, then additional information is needed. It would need to be demonstrated that: (i) NOAELs in the two animal species are defined by local toxicity (e.g., gastrointestinal) that directly scale by body weight across species, (ii) other drugs in the class exhibit similar toxicity in humans and animals at doses that correlate across species on an mg/kg basis, (iii) other pharmacological (e.g., pharmacologically active dose) and toxicological (e.g., MTD) endpoints also scale by mg/kg across species, or (iv) there is good correlation between plasma exposure (Cmax and AUC) and mg/kg dose among species.
Since this approach results in a higher MRSD than when dose is normalized by BSA, it should be used with caution. Published studies indicate that such an approach has led, in a number of cases, to most unfortunate results (including death), at least in animals (25,26). Based on their examination of data from Freireich et al. (9) and Schein et al. (10), Watanabe et al. (11) concluded that interspecies scaling based on mg/kg would have overpredicted the MTD in humans in every case examined.
Was this article helpful?