Coupling biorelevant dissolution methods with physiologically based pharmacokinetic modelling to forecast in-vivo performance of solid oral dosage forms.

OBJECTIVES: To summarize the basis for and progress with the development of in-vitro-in-silico-in-vivo (IV-IS-IV) relationships for oral dosage forms using physiologically based pharmacokinetic (PBPK) modelling, with the focus on predicting the performance of solid oral dosage forms in humans. KEY FINDINGS: Various approaches to forecasting oral absorption have been reported to date. These range from simple dissolution tests, through biorelevant dissolution testing and laboratory simulations of the gastrointestinal (GI) tract, to the use of PBPK modelling to predict oral drug absorption based on the physicochemical parameters of the drug substance. Although each of these approaches can be useful for qualitative predictions, forecasting oral absorption on a quantitative basis with an individual approach is only possible for selected drug/dosage form combinations. By integrating biorelevant dissolution test results with the PBPK models, it has become possible to achieve quantitatively accurate as well as qualitative predictions of plasma profiles after oral dosing for both immediate and modified release formulations. SUMMARY: With further refinement of both the biorelevant dissolution testing methods and the PBPK models, it should be possible to expedite the development and regulatory approval of optimized dosage forms and dosing conditions.

Precipitation in the small intestine may play a more important role in the in vivo performance of poorly soluble weak bases in the fasted state: case example nelfinavir.

The aim of this study was to evaluate the utility of biorelevant dissolution tests coupled with in silico simulation technology to forecast in vivo bioperformance of poorly water-soluble bases, using nelfinavir mesylate as a model compound. An in silico physiologically based pharmacokinetic (PBPK) model for poorly water-soluble, weakly basic drugs was used to generate plasma profiles of nelfinavir by coupling dissolution results and estimates of precipitation with standard gastrointestinal (GI) parameters and the disposition pharmacokinetics of nelfinavir. In vitro dissolution of nelfinavir mesylate film-coated tablets was measured in biorelevant and compendial media. Drug precipitation in the small intestine was estimated from crystal growth theory. GI parameters (gastric emptying rate and fluid volume) appropriate to the dosing conditions (fasting and fed states) were used in the PBPK model. The disposition parameters of nelfinavir were estimated by fitting compartmental models to the in vivo oral PK data. The in vivo performance in each prandial state was simulated with the PBPK model, and predicted values for AUC and C(max) were compared to observed values. Dissolution results in FaSSIF-V2 and FeSSIF-V2, simulating the fasting and fed small intestinal conditions, respectively, correctly predicted that there would be a positive food effect for nelfinavir mesylate, but overestimated the food effect observed in healthy human volunteers. In order to better predict the food effect, an in silico PBPK simulation model using STELLA® software was evolved. Results with the model indicated that invoking drug precipitation in the small intestine is necessary to describe the in vivo performance of nelfinavir mesylate in the fasted state, whereas a good prediction under fed state conditions is obtained without assuming any precipitation. In vitro-in silico-in vivo relationships (IVISIV-R) may thus be a helpful tool in understanding the critical parameters that affect the oral absorption of poorly soluble weak bases.

Forecasting in vivo oral absorption and food effect of micronized and nanosized aprepitant formulations in humans.

This study coupled results from biorelevant dissolution tests with in silico simulation technology to forecast in vivo oral absorption of micronized and nanosized aprepitant formulations in the pre- and post-prandial states. In vitro dissolution tests of the nanosized aprepitant formulation and micronized drug were performed in biorelevant and compendial media. An in silico physiologically based pharmacokinetic (PBPK) model was developed based on STELLA software using dissolution kinetics, standard gastrointestinal (GI) parameters and post-absorptive disposition parameters. GI parameters (gastric emptying rate and GI fluid volume) were varied according to the dosing conditions. Disposition parameters were estimated by fitting compartmental models to the in vivo oral PK data. Predictions of in vivo performance in each prandial state were evaluated using the AUC and C(max) generated from the simulated PK profiles. To predict oral absorption from the extremely fast dissolving nanosized aprepitant formulation, several variations on a previously published model were evaluated. Although models that assumed that the formulation behaved as an oral solution or that adjusted the dissolution kinetics according to the different numbers of particles per gram between micronized and nanosized aprepitant generated profiles similar to the observed in vivo data in the fed state, simulated profiles for the fasted state showed much faster absorption than that observed in the in vivo data. This appeared to result from the assumption of no absorption restrictions in those models. To better predict in vivo performance in both fasted and fed states, a model that adds permeability restrictions to absorption was applied. This model not only simulated the in vivo profiles for aprepitant well in both prandial states, but also predicted the dependency of the pharmacokinetics on the dose and the particle size of aprepitant. In conclusion, a model based on STELLA software combined with dissolution results in biorelevant media successfully forecasts the in vivo performance of both nanosized and micronized formulations of aprepitant in the fed and fasted states. Although dissolution is the primary limitation to the rate of absorption for micronized aprepitant, some permeability restrictions are revealed for the nanosized formulation. The results also indicate that biorelevant dissolution media have strong advantages over compendial media in forecasting the in vivo behavior of aprepitant.

Prediction of food effects on the absorption of celecoxib based on biorelevant dissolution testing coupled with physiologically based pharmacokinetic modeling.

Since the rate-determining step to the intestinal absorption of poorly soluble drugs is the dissolution in the gastrointestinal (GI) tract, postprandial changes in GI physiology, in addition to any specific interactions between drug and food, are expected to affect the pharmacokinetics and bioavailability of such drugs. In this study, in vitro dissolution testing using biorelevant media coupled with in silico physiologically based pharmacokinetic (PBPK) modeling was applied to the prediction of food effects on the absorption of a poorly soluble drug, celecoxib, from 200mg capsules. A PBPK model was developed based on STELLA software using dissolution kinetics, solubility, standard GI parameters and post-absorptive disposition parameters. Solubility, dissolution profiles and initial dissolution rate from celecoxib 200mg capsules were measured in biorelevant and compendial media. Standard GI parameters (gastric emptying rate and fluid volume) were varied according to the dosing conditions. Disposition parameters were estimated by fitting compartmental models to the oral PK data, since intravenous data are not available for celecoxib. Predictions of food effects and average plasma profiles were evaluated using the AUC and C(max) and the difference factor (f(1)). An approximately 7-fold difference in the maximum percentage dissolved was observed in in vitro dissolution tests designed to represent the fed and fasted states. By contrast, the food effect estimated by simulating the plasma profiles with the PBPK model predicted only a slight delay in the peak plasma level ( approximately 1h), and modest increases in the C(max) and AUC of approximately 1.9-fold and 1.3-fold in the fed state, respectively. The PBPK approach, combining in silico simulation coupled with biorelevant dissolution test results, thus corresponds much better to the food effect observed for celecoxib in vivo. Additionally, point estimates of AUC and C(max) as well as f(1) calculations demonstrated clear advantages of using results in biorelevant rather than compendial media in the PBPK model.

Modulation of intestinal P-glycoprotein function by cremophor EL and other surfactants by an in vitro diffusion chamber method using the isolated rat intestinal membranes.

Effects of various surfactants on the transport of rhodamine123, a P-glycoprotein (P-gp) substrate, across the isolated rat intestinal membranes were examined by an in vitro diffusion chamber system. The jejunal serosal-to-mucosal transport (Jsm) of rhodamine123 was more than threefold greater than its mucosal-to-serosal transport (Jms), suggesting that the net movement of rhodamine123 across the rat jejunum was preferentially secretory direction. There exists a regional difference in the intestinal transport of rhodamine123 and the secretory directed transport was remarkably observed in the jejunum. The Jsm/Jms ratio of rhodamine123 decreased in the presence of 0.3 mM verapamil and 10 mM sodium azide (NaN3) + 1 mM sodium fluoride (NaF), confirming that rhodamine123 might be secreted from the intestinal tissue into the lumen by a P-gp-mediated efflux system. Nonionic surfactants [0.1% Cremophor EL, Tween 80 and n-dodecyl-beta-D-maltopyranoside (LM)] reduced the Jsm/Jms ratio of rhodamine123, whereas its ratio was not influenced in the presence of 0.1% cationic surfactant (hexadecyltrimethylammonium bromide, C16TAB) and anionic surfactant (sodium dodecyl sulfate, SDS). Therefore, these findings suggested that charge of surfactants was possibly related to the action of these surfactants on the intestinal absorption of P-gp substrates. On the other hand, the transfer of rhodamine123 was not affected by the addition of Cremophor EL to the serosal side. Because the c.m.c. of Cremophor EL is 0.0095 w/v%, interactions between rhodamine123 and the micellar form of Cremophor EL may decrease the P-gp-mediated efflux of rhodamine123 at higher concentrations. In the kinetic analysis, the Vmax value (nmol/min/g wet tissue) of rhodamine123 decreased, although the Km value (mM) was constant in the presence of Cremophor EL. Therefore, Cremophor EL inhibited the efflux transport of rhodamine123 in a noncompetitive manner. Cremophor EL did not affect the transport of [14C]Gly-Sar and [3H]3-O-methyl-D-glucose, suggesting that the action of Cremophor EL might be P-gp specific. These findings indicated that nonionic surfactants including Cremophor EL and Tween 80 may be useful pharmaceutical excipients for inhibiting the function of P-gp, thereby increasing the intestinal absorption of various drugs, which are secreted by a P-gp-mediated efflux system in the intestine.

A novel method for estimation of transition temperature for polymorphic pairs in pharmaceuticals using heat of solution and solubility data.

A novel method for thermodynamic stability studies of polymorphic drug substances has been developed. In order to estimate the transition temperature for an enantiotropic polymorphic pair, a formula for calculating the temperature at which the solubilities of each polymorph become equal has been derived with heat of solution and solubility as the variables. This formula is based on the assumption that van't Hoff plots (logarithmic solubility versus reciprocal of absolute temperature plots) of each polymorph show a straight line (heat of solution is independent of temperature) whose slope can be expressed as a function of heat of solution. The transition temperatures for seratrodast, acetazolamide and carbamazepine polymorphic pairs calculated by the formula were in good agreement with the results of previous studies. Furthermore, the calculated transition temperature for the indomethacin polymorphic pair was above the melting point, an unrealistic temperature range, suggesting that these polymorphs are monotropically related. Since this formula requires solubility data at only one arbitrary temperature other than heat of solution data for both polymorphs in a polymorphic pair, the proposed method is much faster than the conventional method requiring solubility data at five or more different temperatures for the preparation of van't Hoff plots.