Re: QSAR - QSAR: Published correlations of animal versus human PK data for numerouscompounds ?

Jim,

For bioavailabilities, try W.E. Sietsema, Intl. J. Clin. Pharmacol. Therapy
Toxicol. 1989, 27, 179-211. He has a good compilation of carefully curated data
showing that he correlation here is pretty poor in general - best for rats, as I
recall. More importantly, he discusses the underlying physiological differences
that account for the discrepancy.

It seems to me that for bioavailability, animal data is useful for validating
methods for constructing models, but generally not for training models intended
for predicting human PK behavior.

Bob C.

james.metz[A]abbott.com wrote:

> QSAR Society,
>
> Numerous publications exist which attempt to develop QSAR models
> to predict ADMET in humans
> and other preclinical animal species. A significant component of the
> pharmaceutical industry presupposes that
> preclinical pharmacokinetics (PK) in animals is at least *somewhat*
> predictive of PK in humans.
>
> Is anyone aware of any publications which clearly show the
> correlation (or lack thereof !) between experimental
> PK data in animals and humans for more than a few compounds? For example,
> is anyone aware of any publications
> showing a plot of po %F in rats (or mice) versus po %F humans for say 20
> - 30 (hopefully more) compounds?
>
> I am aware of publications with a few isolated comparisons.
> However, I am looking for a more definitive,
> "comprehensive" analysis. I have yet to find any !

Received on 2003-10-16 - 12:49 GMT