Predicting Human Bioavailability

Bioavailability is a measure of the extent of a drug reaching the systemic circulation from its point of administration. From an oral dose therefore, it depends both on the degree of absorption and first pass clearance experienced by the dose. Assuming good estimates of absorption and predicted clearance are available, the bioavailability (F) can be predicted using eqn (13.11). 

---

Obach et al. [27] proposed a model to predict human bioavailability from a retrospective study of in vitro metabolism and in vivo animal pharmacokinetic (PK) data. While their model yielded acceptable predictions (within a factor of 2) for an expansive group of compounds, it relied extensively on in vivo animal PK data for interspecies scaling in order to estimate human PK parameters. Animal data are more time-consuming and costly to obtain than are permeability and metabolic clearance data hence, this approach may be limited to the later stages of discovery support when the numbers of compounds being evaluated are fewer. 

L. X., Predicting human oral bioavailability of a compound development of a novel quantitative structure-bioavailability relationship, Pharm. 

As previously noted, the bioavailability of a compound is a complex function that includes contributions from absorption and clearance. Since the molecule must undergo these biological processes in all species, there is a temptation to assume a relationship between the bioavailability between species, and hence that human bioavailability can be predicted by such relationships. Although a linear correlation has been demonstrated for the rate/extent of absorption (% oral dose absorbed) between species for various drugs [6-8], there is clearly a lack of correlation for bioavailability between species [2, 8]. Figure 19.1 shows the excellent correlation in... 

In the first example, the predicted oral absorption for a series of ACE inhibitors has been compared with published values of human bioavailability. For the generation of calculated absorption, a sigmoidal curve between observed human absorption and PSA for a series of reference compounds was used [25], The predicted oral absorption for ACE inhibitors is plotted against the calculated PSA values is shown in Fig. 19.6 however, as expected, only a partial correlation existed between predicted absorption and observed in vivo bioavailability. 

In vitro Model for Predicting Oral Bioavailability in Human and other Species... 

Fig. 19.8. Prediction of human bioavailability from calculated human absorption using polar surface area (PSA) for a series of calcium antagonists [25],...

Andrews CW, Bennett L and Yu LX (2000) Predicting Human Oral Bioavailability of a Compound Development of a Novel Quantitative Structure-Bioavailability Relationship. Pharm Res 17 pp 639-644. 

Recently, Gasteiger et al. [59] reported several models to predict human oral bioavailability using Hou and Wang s data set. A set of ADRIANA.Code and Cerius2 descriptors were calculated, and MLR analysis was performed. The best linear model had r2 of 0.18 and RMSD of 31.15. When a set of subsets was cherry-picked so that each subset had either a common functional group or a similar pharmacological activity, the r2 values were improved and RMSD values dropped. But the performance of those models was still not satisfactory the standard errors were above 20.0 and r2 was lower than 0.6. 

Andrews, C.W., Bennett, L., Yu, L.X. Predicting human oral bioavailability of a compound development of a novel quantitative structure-bioavailability relationship. Pharm. Res. 2000, 17, 639 4. 

Hou et al. compiled a database of human bioavailability for 768 compounds, which is publicly available [34], These authors used a cutoff of 20% as acceptable. This can be questioned as F% up to about 30-40% can show considerable interindividual variability. It was concluded that h % of highly metabolized compounds cannot be well predicted from simple molecular descriptors as these do not encode for metabolism. 

A recent review gives a comprehensive survey of the state of the art in modeling human bioavailability [38]. Commercial software for the prediction of bioavailability using QSAR approaches include ADME Boxes [www.ap-algorithms.com], truPK [trupk.strandgenomics.com], and KnowItAll [www.knowittallcom]. 

---