Category: Preclinical Development
Purpose: Oral absorption of poorly water-soluble drugs saturates at the high dose range due to the saturation in solubility in the GI tract. This nonlinearity in oral absorption often caused the insufficient systemic exposure of orally administered drugs in a dose escalation study which resulted in the failure in toxicity assay. To streamline the process of oral drug development, therefore, assessment of the linearity in oral absorption is required for early study with small animals (such as rats), later study with big animals (such as dogs) as well as a clinical study. In order to understand the factors causing a saturation in oral drug absorption, concept of Maximum Absorbable Dose (MAD) is often referred and described as
MAD = Cs,free x Peff x SA x Tsi
where Cs,free, Peff, SA, and Tsi represent the saturated solubility of free drug in the intestinal fluid, effective permeability to the intestinal membrane, effective surface area and transit time in the GI tract, respectively. In the MAD equation, Cs,free and Peff are physicochemical parameters of the drug, while SA and Tsi are physiological parameters of the GI tract which depend on the animal species. In this study, based on the concept of MAD, we have tried to predict the linearity of oral drug absorption in humans from the results of in vitro dissolution study and in vivo absorption study in rats. As a parameter to express the linearity in absorption, maximum linear dose (MLD) which represents a maximum oral dose giving a linear absorption, thus the constant Fa (fraction dose absorbed), was newly defined.
Methods: Physiological parameters of the GI tract in rat and human, SA/V (surface area/volume ratio of the intestinal tract) and Tsi, were obtained from the literatures1,2). Griseofulvin was used as a model of poorly soluble drug. Solubility of griseofulvin in biorelevant medium was measured by in vitro dissolution study. Human and rat fasted state simulated intestinal fluid (FaSSIFhuman, FaSSIFrat) were prepared with 3 mM bile acid and 0.75 mM lecithin for FaSSIFhuman and 50 mM bile acid and 3.7 mM lecithin for FaSSIFrat at pH 6.5. MLD value of griseofulvin in rat (MLDrat) was determined by oral administration of griseofulvin as a suspension at various doses. Area under the plasma concentration time curve (AUC) was calculated for each dose, then AUCmax was estimated by fitting dose-AUC profiles to non-linear least square equation. Then, the doses giving 50% and 80% of AUCmax value (MLDrat,50 and MLDrat,80) were obtained and used as index of the linear dose range after oral administration. Finally, MLDhuman was predicted by combining physiological parameters of GI tract with the results in in vitro dissolution and in vivo rat studies. On the other hand, the clinical data of oral administration of griseofulvin were extracted from the reports3,4) and, from the relation between oral dose and AUC, MLDhuman value was calculated. Observed and predicted values MLDhuman were compared to validate our approach to predict the linear dose range in human.
Results: The SA/V x Tsi values in human and rat were reported to be 8.05, and 25.5 (h/cm), respectively. Solubility of griseofulvin in FaSSIFhuman and FaSSIFrat were 14.0, and 97.5 (μg/mL), respectively, giving the ratio (FaSSIFrat / FaSSIFhuman) about 6.9. In the in vivo rat absorption study, MLDrat,50 and MLDrat,80 values were calculated to be 20.0, and 80.1 (mg), respectively, which indicated that solubility of griseofulvin in rat GI tract was saturating at these dose range. The linear dose range in human was predicted by using the ratios of SA/V x Tsi and solubility, and MLDrat value. Predicted MLDhuman,50 and MLDhuman,80 values were about 230 mg, and 910 mg, respectively. These values were in good agreement with the observed MLD values, 220 mg as MLDhuman,50 and 870 mg as MLDhuman,80.
Conclusion: In conclusion, we have successfully demonstrated that our approach in this study to predict the MLD value in human was appropriate and is useful for planning a dose-escalation study in drug development. In a future project, MLDhuman values of various drugs will be evaluated using the same protocol with that established in the present study.
1) Int J Pharm. 2011 405 (1-2) 79-89
2) Food Chem Toxicol. 2001 39(3) 209-228
3) J Pharm Sci. 1982 71 (10) 1165-1169
4) J Pharmacokinet Biopharm. 1980 8(4) 347-362