relationship between Fa and Caco-2 cell perme-
ability, expressed as the apparent permeability
constant (Papp), as follows:
ifwPapp< 10 ^6 cms^1 ;then Fa¼ 0 20%
ifw 1 Papp 10 10 ^6 cms^1 ; then Fa¼ 20 70%
ifwPapp> 10 ^5 cms^1 then Fa70%
The use of Caco-2 cell permeability studies has
resulted in more accurate oral bioavailability pre-
dictions. Using the predicted hepatic clearance for
compound X in humans (see above), estimatingFa
by extrapolation from the Caco-2 cellPappand
assuming hepatic blood flow for humans (see, for
example Raneet al., 1977) of 20 ml min^1 kg^1 ,
the human oral bioavailability of 69–98% is pre-
dicted for compound X. This compares well with
the known oral bioavailability of this compound in
rats and dogs (83 and 72%, respectively).
8.2 Prediction from animals
to humansin vivo
Elementary aspects
Allometric scaling is an empirical method for
predicting physiological, anatomical and pharma-
cokinetic measures across species in relation
to time and size (Boxenbaum, 1982; Ings, 1990;
Boxenbaum and DiLea, 1995). Allometric scaling
is based on similarities among species in their
physiology, anatomy and biochemistry, coupled
with the observation that smaller animals perform
physiological functions that are similar to larger
animals, but at a faster rate. The allometric equa-
tion isY¼aWb, and a log transformation of this
formula yields the straight line:
logY¼blogWþloga;
whereYis the pharmacokinetic or physiological
variable of interest,ais the allometric coefficient
(and logais the intercept of the line),Wis the body
weight andbis the allometric exponent (slope of
the line).
One of the first applications of allometric scaling
was the use of the toxicity of anticancer agents in
animals to predict toxicity in humans children. It
was observed that the toxic dose of a drug is similar
among species when the dose is compared on the
basis of body surface area (Freireichet al., 1966).
For most vertebrate species, the body weight/
volume ratio varies very little, but the surface
area/volume ratio increases as species become
smaller. Allometric correction of dose multiples
in toxicology (compared with proposed human
doses) is thus important, especially when small
rodents provide the principal toxicology coverage.
Body surface area (Y) is related to body weight
(W, in kg) by the formula:
Y¼ 0 : 1 W^0 :^67
This allometric relationship between body surface
area and species body weight then allows for a
simple conversion of drug doses across species
(Figure 8.3), and allometrically equivalent doses
of drugs (mg kg^1 ) can be calculated for any
species (Table 8.4). The conversion factor (km) is
simply the body weight divided by the body sur-
face area. Thus, using the km factors, the dose in
Species 1 (in mg kg^1 ) is equivalent to (kmspecies2/
kmspecies1) times the dose in Species 2 (in mg
kg^1 ). For example, a 50 mg kg^1 dose of drug
in mouse would be equivalent to a 4.1 mg kg^1
dose in humans, that is approximately one-twelfth
of the dose (Table 8.4). Likewise, the conversion
factor can be used to calculate equivalent doses
between any species. An equivalent dose in milli-
gram per kilogram in rat would be twice that for
mouse.
Allometric approaches to drug discovery
Using limited data, allometric scaling may be use-
ful in drug discovery. We assume that, for the
formulaY¼aWb, the value of the power function
‘b’ (or slope of the line from a log vs. log plot) is
drug independent, unlike the intercept ‘a’, which is
drug dependent. By doing this, we can use data
from a single species (rat) to successfully predict
the pharmacokinetics of compound X in humans
86 CH8 PHASE I: THE FIRST OPPORTUNITY FOR EXTRAPOLATION FROM ANIMAL DATA