While avian cytochrome P450 activities have been reported, their expression and
role in avian drug metabolism are not well documented (Walker 1998 ). On the basis
of published data (Hunter et al. 2008 ), it is evident that allometric scaling between
mammalian and avian species produces highly biased estimates of drug clearance,
irrespective of whether the drug is cleared in mammalian species primarily through
renal or hepatic mechanisms.
Choice of the species to be included in the data analysis is another potential
source of error. It seems likely that inclusion of at least one large animal in the
scaling procedure can improve predictions (Mahmood et al. 2006 ; Table 3 ). In large
animals, correction factors (used in human first-dose determination) could not be
applied because there was no correlation between the exponents of the simple
allometry and the correction factor used to improve dose predictions (Mahmood
et al. 2006 ). In addition, the vast majority of large animals for which pharmacoki-
netic data are available are herbivores (cattle, horse, and elephant). Conversely,
smaller animal species are generally either omnivores (mouse, rat, monkey, and
humans) or carnivores (dogs and cats). Differences in diet can influence both drug
metabolism and renal elimination (Martinez 2005 ). Ideally, the species selected
would be similar in all the major physiological functions and differ only in size.
Unfortunately, there is no method to predict, a priori, which animal species will be
best suited for inclusion in the interspecies predictions for a particular species.
Although data for large carnivorous species (such as lion, tiger, and polar bear)
might be helpful for improving interspecies predictions, the obvious dangers hinder
progress in this area of research.
3 Discussion and Conclusions
Drug–dose scaling of therapeutic agents is of particular interest to veterinarians and
pharmacokineticists because this method can, when used appropriately, provide
an estimate for designing therapeutic dosage regimens in an unstudied species.
Sedgwick’s metabolic scaling approach implies that body metabolic rate can be
applied to almost any drug and any species derive a therapeutic dosage regimen
(Dorrestein 2000 ; Sedgwick 1993 ; Sedgwick and Borkowski 1996 ). Although this
would be very useful clinically, it oversimplifies the relationship between basal
metabolic rate and drug pharmacokinetics. Sedgwick’s method requires only that
basal metabolic rate of a control species, together with a known therapeutic dose in
that species, be used to calculate the dose for the unknown species for which the
metabolic rate can be estimated from its weight (Sedgwick 1993 ). However, on
the basis of the published literature (Careau et al. 2007 ; Jacobson 1996 ; Packard
and Birchard 2008 ) Sedgwick’s approach does not appear to be appropriate. This
is confirmed by the work of Mahmood et al. ( 2006 ); Martinez et al. ( 2006 ); and
Hunter et al. ( 2008 ).
These observations emphasize the importance of applying fundamental princi-
ples associated with allometric analysis and using appropriate caution when
Interspecies Allometric Scaling 149