In Silico Prediction of Human Clinical Pharmacokinetics with ANDROMEDA by Prosilico: Predictions for an Established Benchmarking Data Set, a Modern Small Drug Data Set, and a Comparison with Laboratory Methods.

There is an ongoing aim to replace animal and in vitro laboratory models with in silico methods. Such replacement requires the successful validation and comparably good performance of the alternative methods. We have developed an in silico prediction system for human clinical pharmacokinetics, based on machine learning, conformal prediction and a new physiologically-based pharmacokinetic model, i.e. ANDROMEDA. The objectives of this study were: a) to evaluate how well ANDROMEDA predicts the human clinical pharmacokinetics of a previously proposed benchmarking data set comprising 24 physicochemically diverse drugs and 28 small drug molecules new to the market in 2021; b) to compare its predictive performance with that of laboratory methods; and c) to investigate and describe the pharmacokinetic characteristics of the modern drugs. Median and maximum prediction errors for the selected major parameters were ca 1.2 to 2.5-fold and 16-fold for both data sets, respectively. Prediction accuracy was on par with, or better than, the best laboratory-based prediction methods (superior performance for a vast majority of the comparisons), and the prediction range was considerably broader. The modern drugs have higher average molecular weight than those in the benchmarking set from 15 years earlier (ca 200 g/mol higher), and were predicted to (generally) have relatively complex pharmacokinetics, including permeability and dissolution limitations and significant renal, biliary and/or gut-wall elimination. In conclusion, the results were overall better than those obtained with laboratory methods, and thus serve to further validate the ANDROMEDA in silico system for the prediction of human clinical pharmacokinetics of modern and physicochemically diverse drugs.

The Impact of Reference Data Selection for the Prediction Accuracy of Intrinsic Hepatic Metabolic Clearance.

In vitro-in vivo prediction results for hepatic metabolic clearance (CLH) and intrinsic CLH (CLint) vary widely among studies. Reasons are not fully investigated and understood. The possibility to select favorable reference data for in vivo CLH and CLint and unbound fraction in plasma (fu) is among possible explanations. The main objective was to investigate how reference data selection influences log in vitro and in vivo CLint-correlations (r2). Another aim was to make a head-to-head comparison vs an in silico prediction method. Human hepatocyte CLint-data for 15 compounds from two studies were selected. These were correlated to in vivo CLint estimated using different reported CLH- and fu-estimates. Depending on the choice of reference data, r2 from two studies were 0.07 to 0.86 and 0.06 to 0.79. When using average reference estimates a r2 of 0.62 was achieved. Inclusion of two outliers in one of the studies resulted in a r2 of 0.38, which was lower than the predictive accuracy (q2) for the in silico method (0.48). In conclusion, the selection of reference data appears to play a major role for demonstrated predictions and the in silico method showed higher accuracy and wider range than hepatocytes for human in vivo CLint-predictions.

In Silico Predictions of the Gastrointestinal Uptake of Macrocycles in Man Using Conformal Prediction Methodology.

The gastrointestinal uptake of macrocyclic compounds is not fully understood. Here we applied our previously validated integrated system based on machine learning and conformal prediction to predict the passive fraction absorbed (fa), maximum fraction dissolved (fdiss), substrate specificities for major efflux transporters and total fraction absorbed (fa,tot) for a selected set of designed macrocyclic compounds (n = 37; MW 407-889 g/mol) and macrocyclic drugs (n = 16; MW 734-1203 g/mole) in vivo in man. Major aims were to increase the understanding of oral absorption of macrocycles and further validate our methodology. We predicted designed macrocycles to have high fa and low to high fdiss and fa,tot, and average estimates were higher than for the larger macrocyclic drugs. With few exceptions, compounds were predicted to be effluxed and well absorbed. A 2-fold median prediction error for fa,tot was achieved for macrocycles (validation set). Advantages with our methodology include that it enables predictions for macrocycles with low permeability, Caco-2 recovery and solubility (BCS IV), and provides prediction intervals and guides optimization of absorption. The understanding of oral absorption of macrocycles was increased and the methodology was validated for prediction of the uptake of macrocycles in man.

In silico predictions of the human pharmacokinetics/toxicokinetics of 65 chemicals from various classes using conformal prediction methodology.

Pharmacokinetic/toxicokinetic (PK/TK) information for chemicals in humans is generally lacking. Here we applied machine learning, conformal prediction and a new physiologically-based PK/TK model for prediction of the human PK/TK of 65 chemicals from different classes, including carcinogens, food constituents and preservatives, vitamins, sweeteners, dyes and colours, pesticides, alternative medicines, flame retardants, psychoactive drugs, dioxins, poisons, UV-absorbents, surfactants, solvents and cosmetics.About 80% of the main human PK/TK (fraction absorbed, oral bioavailability, half-life, unbound fraction in plasma, clearance, volume of distribution, fraction excreted) for the selected chemicals was missing in the literature. This information was now added (from in silico predictions). Median and mean prediction errors for these parameters were 1.3- to 2.7-fold and 1.4- to 4.8-fold, respectively. In total, 59 and 86% of predictions had errors <2- and <5-fold, respectively. Predicted and observed PK/TK for the chemicals was generally within the range for pharmaceutical drugs.The results validated the new integrated system for prediction of the human PK/TK for different chemicals and added important missing information. No general difference in PK/TK-characteristics was found between the selected chemicals and pharmaceutical drugs.

In silico prediction of volume of distribution of drugs in man using conformal prediction performs on par with animal data-based models.

Volume of distribution at steady state (Vss) is an important pharmacokinetic endpoint. In this study we apply machine learning and conformal prediction for human Vss prediction, and make a head-to-head comparison with rat-to-man scaling, allometric scaling and the Rodgers-Lukova method on combined in silico and in vitro data, using a test set of 105 compounds with experimentally observed Vss.The mean prediction error and % with <2-fold prediction error for our method were 2.4-fold and 64%, respectively. 69% of test compounds had an observed Vss within the prediction interval at a 70% confidence level. In comparison, 2.2-, 2.9- and 3.1-fold mean errors and 69, 64 and 61% of predictions with <2-fold error was reached with rat-to-man and allometric scaling and Rodgers-Lukova method, respectively.We conclude that our method has theoretically proven validity that was empirically confirmed, and showing predictive accuracy on par with animal models and superior to an alternative widely used in silico-based method. The option for the user to select the level of confidence in predictions offers better guidance on how to optimise Vss in drug discovery applications.

Comparison between lab variability and in silico prediction errors for the unbound fraction of drugs in human plasma.

Variability of the unbound fraction in plasma (fu) between labs, methods and conditions is known to exist. Variability and uncertainty of this parameter influence predictions of the overall pharmacokinetics of drug candidates and might jeopardise safety in early clinical trials. Objectives of this study were to evaluate the variability of human in vitro fu-estimates between labs for a range of different drugs, and to develop and validate an in silico fu-prediction method and compare the results to the lab variability.A new in silico method with prediction accuracy (Q2) of 0.69 for log fu was developed. The median and maximum prediction errors were 1.9- and 92-fold, respectively. Corresponding estimates for lab variability (ratio between max and min fu for each compound) were 2.0- and 185-fold, respectively. Greater than 10-fold lab variability was found for 14 of 117 selected compounds.Comparisons demonstrate that in silico predictions were about as reliable as lab estimates when these have been generated during different conditions. Results propose that the new validated in silico prediction method is valuable not only for predictions at the drug design stage, but also for reducing uncertainties of fu-estimations and improving safety of drug candidates entering the clinical phase.

Advances in Predictions of Oral Bioavailability of Candidate Drugs in Man with New Machine Learning Methodology.

Oral bioavailability (F) is an essential determinant for the systemic exposure and dosing regimens of drug candidates. F is determined by numerous processes, and computational predictions of human estimates have so far shown limited results. We describe a new methodology where F in humans is predicted directly from chemical structure using an integrated strategy combining 9 machine learning models, 3 sets of structural alerts, and 2 physiologically-based pharmacokinetic models. We evaluate the model on a benchmark dataset consisting of 184 compounds, obtaining a predictive accuracy (Q2) of 0.50, which is successful according to a pharmaceutical industry proposal. Twenty-seven compounds were found (beforehand) to be outside the main applicability domain for the model. We compare our results with interspecies correlations (rat, mouse and dog vs. human) using the same dataset, where animal vs. human-correlations (R2) were found to be 0.21 to 0.40 and maximum prediction errors were smaller than maximum interspecies differences. We conclude that our method has sufficient predictive accuracy to be practically useful with applications in human exposure and dose predictions, compound optimization and decision making, with potential to rationalize drug discovery and development and decrease failures and overexposures in early clinical trials with candidate drugs.

Prediction of in vivo rat biliary drug clearance from an in vitro hepatocyte efflux model.

Well-established techniques are available to predict in vivo hepatic uptake and metabolism from in vitro data, but predictive models for biliary clearance remain elusive. Several studies have verified the expression and activity of ATP-binding cassette (ABC) efflux transporters central to biliary clearance in freshly isolated rat hepatocytes, raising the possibility of predicting biliary clearance from in vitro efflux measurements. In the present study, short-term plated rat hepatocytes were evaluated as a model to predict biliary clearance from in vitro efflux measurements before major changes in transporter expression known to take place in long-term hepatocyte cultures. The short-term cultures were carefully characterized for their uptake and metabolic properties using a set of model compounds. In vitro efflux was studied using digoxin, fexofenadine, napsagatran, and rosuvastatin, representing compounds with over 100-fold differences in efflux rates in vitro and 60-fold difference in measured in vivo biliary clearance. The predicted biliary clearances from short-term plated rat hepatocytes were within 2-fold of measured in vivo values. As in vitro efflux includes both basolateral and canalicular effluxes, pronounced basolateral efflux may introduce errors in predictions for some compounds. In addition, in vitro rat hepatocyte uptake rates corrected for simultaneous efflux predicted rat in vivo hepatic clearance of the biliary cleared compounds with less than 2-fold error. Short-term plated hepatocytes could thus be used to quantify hepatocyte uptake, metabolism, and efflux of compounds and considerably improve the prediction of hepatic clearance, especially for compounds with a large biliary clearance component.

In silico categorization of in vivo intrinsic clearance using machine learning.

Machine learning has recently become popular and much used within the life science research domain, e.g., for finding quantitative structure-activity relationships (QSARs) between molecular structures and different biological end points. In the work presented here, we have applied orthogonal partial least-squares (OPLS), principal component analysis (PCA), and random forests (RF) methods for classification as well as regression analysis to a publicly available in vivo data set in order to assess the intrinsic metabolic clearance (CL(int)) in humans. The derived classification models are able to identify compounds with CL(int) lower and higher than 1500 mL/min, respectively, with nearly 80% accuracy. The most relevant descriptors are of lipophilicity and charge/polarizability types. Furthermore, the accuracy from a classification model based on regression analysis, using the 1500 mL/min cutoff, is also around 80%. These results suggest the usefulness of machine learning techniques to derive robust and predictive models in the area of in vivo ADMET (absorption, distribution, metabolism, elimination, and toxicity) modeling.

The volume of distribution is an indicator of poor in vitro-in vivo extrapolation of clearance for acidic drugs in the rat.

Abstract 1. We applied the regression offset approach to predict rat in vivo intrinsic clearance (CLint) for 54 new chemical acid entities with high plasma protein binding values and low renal clearance (CL). The prediction success was correlated to volume of distribution (Vd), molecular weight (Mw) and CL. 2. A correlation between poor in vitro-in vivo extrapolation (IVIVE) and Vd values distinct from the Vd of albumin (0.1-0.2 L/kg) was revealed. For compounds with a Vd value above 0.5 L/kg, 0% of the predictions of in vivo CLint was within twofold of the observed value, compared to 69% for compounds with a Vd value below 0.5 L/kg. 3. Compounds with a Mw below 450 g/mol demonstrated more accurate in vivo CLint predictions than compounds with a Mw above 450 g/mol, i.e. 63% compared to 21% within twofold. For compounds with in vivo CL below 30% of the liver blood flow (LBF), 53% of the predictions was within twofold of the observed value, compared to 0% for compounds with CL above 30% of the LBF. 4. We show that accurate IVIVE for acidic compounds with high plasma protein binding and low renal CL can be associated with a low Vd (i.e. around the Vd of albumin) and with a low in vivo CL, and that Mw is an important optimization parameter for pharmacokinetic. This study also further demonstrates the advantages of the application of the regression method for identifying cases when metabolic CL is not the single major elimination pathway.

Prediction of human pharmacokinetics-biliary and intestinal clearance and enterohepatic circulation.

The main objective was to evaluate and propose methods for predicting biliary clearance (CL(bile)) and enterohepatic circulation (EHC) of intact drugs in man. Another aim was to evaluate to role of intestinal drug secretion and propose a method for prediction of intestinal secretion CL (CL(i)). Animal data poorly predict the CL and CL(bile) of biliary excreted drugs, and the suggested molecular weight threshold for bile excretion as the dominant elimination route does not seem to hold. Active transport, low metabolic intrinsic CL (CL(int)) and, as an approximation, permeability (P(e)) less than that of metoprolol is required for substantial CL(bile) to occur. The typical EHC plasma concentration vs time profile (multiple peaks) is demonstrated for many low metabolic CL(int)-compounds with efflux and moderate to high intestinal P(e) and fraction absorbed. Physiologically-based in-vitro to in-vivo (PB-IVIV) methodology with in-vitro intrinsic CL(bile)-data obtained with sandwich-cultured human hepatocytes has generated 2- and 5-fold underpredictions for two compounds with intermediate to high CL(bile). This is despite not considering the unbound fraction. Possible explanations include low transporter activity and diffusion limitations in the in-vitro experiments. Intestinal reabsorption and EHC were also neglected in these predictions and in-vivo CL(bile) estimations. The sandwich model and these reference data are still very useful. Consideration of an empirical scaling factor and a newly developed approach that accounts for intestinal reabsorption and EHC could potentially lead to improved PB-IVIV predictions of CL(bile). Apparently, no attempts have been made to predict CL(i). Elimination via the intestinal route does not appear to be of great importance for the few compounds with available data, but could be equally as important as bile excretion. Net secretion in-vitro P(e) and newly estimated in-vivo intrinsic CL(i) data for digoxin and rosuvastatin could be useful for approximation of CL(i) of other compounds.

The role of permeability in drug ADME/PK, interactions and toxicity--presentation of a permeability-based classification system (PCS) for prediction of ADME/PK in humans.

PURPOSE: The objective was to establish in vitro passive permeability (Pe) vs in vivo fraction absorbed (fa)-relationships for each passage through the human intestine, liver, renal tubuli and brain, and develop a Pe-based ADME/PK classification system (PCS). MATERIALS AND METHODS: Pe- and intestinal fa-data were taken from an available data set. Hepatic fa was calculated based on extraction ratios of the unbound fraction of drugs (with support from animal in vivo uptake data). Renal fa (reabsorption) was estimated using renal pharmacokinetic data, and brain fa was predicted using animal in vitro and in vivo brain Pe-data. Hepatic and intestinal fa-data were used to predict bile excretion potential. RESULTS: Relationships were established, including predicted curves for bile excretion potential and minimum oral bioavailability, and a 4-Class PCS was developed: I (very high Pe; elimination mainly by metabolism); II (high Pe) and III (intermediate Pe and incomplete fa); IV (low Pe and fa). The system enables assessment of potential drug-drug transport interactions, and drug and metabolite organ trapping. CONCLUSIONS: The PCS and high quality Pe-data (with and without active transport) are believed to be useful for predictions and understanding of ADME/PK, elimination routes, and potential interactions and organ trapping/toxicity in humans.

The highly permeable blood-brain barrier: an evaluation of current opinions about brain uptake capacity.

The blood-brain barrier is often perceived as relatively impermeable, and various cut-off values for zero or limited brain permeability have been suggested. The validity of these values has been evaluated in this review. The barrier appears to have a very high permeability and absorptive capacity: sufficient to absorb compounds with polar surface area >270 A(2), molecular weight > 1,000 Da, log D < -3.5 and equilibrium brain-to-blood concentration ratio <0.01 well. Sufficient intestinal uptake indicates good passive brain uptake potential. The uptake is potentially more sensitive to involvement and changes of active transport than in the intestines. A physiologically based in vitro-in vivo method for prediction of brain uptake is presented.

Prediction of human pharmacokinetics - renal metabolic and excretion clearance.

The kidneys have the capability to both excrete and metabolise drugs. An understanding of mechanisms that determine these processes is required for the prediction of pharmacokinetics, exposures, doses and interactions of candidate drugs. This is particularly important for compounds predicted to have low or negligible non-renal clearance (CL). Clinically significant interactions in drug transport occur mostly in the kidneys. The main objective was to evaluate methods for prediction of excretion and metabolic renal CL (CL(R)) in humans. CL(R) is difficult to predict because of the involvement of bi-directional passive and active tubular transport, differences in uptake capacity, pH and residence time on luminal and blood sides of tubular cells, and limited knowledge about regional tubular residence time, permeability (P(e)) and metabolic capacity. Allometry provides poor predictions of excretion CL(R) because of species differences in unbound fraction, urine pH and active transport. The correlation between fraction excreted unchanged in urine (f(e)) in humans and animals is also poor, except for compounds with high passive P(e) (extensive/complete tubular reabsorption; zero/negligible f(e)) and/or high non-renal CL. Physiologically based in-vitro/in-vivo methods could potentially be useful for predicting CL(R). Filtration could easily be predicted. Prediction of tubular secretion CL requires an in-vitro transport model and establishment of an in-vitro/in-vivo relationship, and does not appear to have been attempted. The relationship between passive P(e) and tubular fraction reabsorbed (f(reabs)) for compounds with and without apparent secretion has recently been established and useful equations and limits for prediction were developed. The suggestion that reabsorption has a lipophilicity cut-off does not seem to hold. Instead, compounds with passive P(e) that is less than or equal to that of atenolol are expected to have negligible passive f(reabs). Compounds with passive P(e) that is equal to or higher than that of carbamazepine are expected to have complete f(reabs). For compounds with intermediate P(e) the relationship is irregular and f(reabs) is difficult to predict. Tubular cells are comparably impermeable (for passive diffusion), and show regional differences in enzymatic and transporter activities. This limits the usefulness of microsome data and makes microsome-based predictions of metabolic CL(R) questionable. Renal concentrations and activities of CYP450s are comparably low, suggesting that CYP450 substrates have negligible metabolic CL(R). The metabolic CL(R) of high-P(e) UDP-glucuronyltransferase substrates could contribute to the total CL.

Prediction of human pharmacokinetics--gut-wall metabolism.

Intestinal mucosal cells operate with different metabolic and transport activity, and not all of them are involved in drug absorption and metabolism. The fraction of these cells involved is dependent on the absorption characteristics of compounds and is difficult to predict (it is probably small). The cells also appear comparably impermeable. This shows a limited applicability of microsome intrinsic clearance (CL(int))-data for prediction of gut-wall metabolism, and the difficulty to predict the gut-wall CL (CL(GW)) and extraction ratio (E(GW)). The objectives of this review were to evaluate determinants and methods for prediction of first-pass and systemic E(GW) and CL(GW) in man, and if required and possible, develop new simple prediction methodology. Animal gut-wall metabolism data do not appear reliable for scaling to man. In general, the systemic CL(GW) is low compared with the hepatic CL. For a moderately extracted CYP3A4-substrate with high permeability, midazolam, the gut-wall/hepatic CL-ratio is only 1/35. This suggests (as a general rule) that systemic CL(GW) can be neglected when predicting the total CL. First-pass E(GW) could be of importance, especially for substrates of CYP3A4 and conjugating enzymes. For several reasons, including those presented above and that blood flow based models are not applicable in the absorptive direction, it seems poorly predicted with available methodology. Prediction errors are large (several-fold on average; maximum approximately 15-fold). A new simple first-pass E(GW)-prediction method that compensates for regional and local differences in absorption and metabolic activity has been developed. It has been based on human cell in-vitro CL(int) and fractional absorption from the small intestine for reference (including verapamil) and test substances, and in-vivo first-pass E(GW)-data for reference substances. First-pass E(GW)-values for CYP3A4-substrates with various degrees of gastrointestinal uptake and CL(int) and a CYP2D6-substrate were well-predicted (negligible errors). More high quality in-vitro CL(int)- and in-vivo E(GW)-data are required for further validation of the method.

Prediction of human pharmacokinetics--improving microsome-based predictions of hepatic metabolic clearance.

Physiologically based methods generally perform poorly in predicting in-vivo hepatic CL (CL(H)) from intrinsic clearance (CL(int)) in microsomes in-vitro and unbound fraction in blood (f(u,bl)). Various strategies to improve the predictability have been developed, and inclusion of an empirical scaling factor (SF) seems to give the best results. This investigation was undertaken to evaluate this methodology and to find ways to improve it further. The work was based on a diverse data set taken from Ito and Houston (2005). Another objective was to evaluate whether rationalization of CL(H) predictions can be made by replacing blood/plasma-concentration ratio (C(bl)/C(pl)) measurements with SFs. There were apparently no or weak correlations between prediction errors and lipophilicity, permeability (compounds with low permeability missing in the data set) and main metabolizing CYP450s. The use of CL(int) class (high/low) and drug class (acid/base/neutral) SFs (the CD-SF method) gives improved and reasonable predictions: 1.3-fold median error (an accurate prediction has a 1-fold error), 76% within 2-fold-error, and a median absolute rank ordering error of 2 for CL(H) (n = 29). This approach is better than the method with a single SF. Mean (P < 0.05) and median errors, fraction within certain error ranges, higher percentage with most accurate predictions, and ranking were all better, and 76% of predictions were more accurate with this new method. Results are particularly good for bases, which generally have higher CL(H) and the potential to be incorrectly selected/rejected as candidate drugs. Reasonable predictions of f(u,bl) can be made from plasma f(u) (f(u,pl)) and empirical blood cell binding SFs (B-SFs; 1 for low f(u,pl) acids; 0.62 for other substances). Mean and median f(u,bl) prediction errors are negligible. The use of the CD-SF method with predicted f(u,bl) (the BCD-SF method) also gives improved and reasonable results (1.4-fold median error; 66% within 2-fold-error; median absolute rank ordering error = 1). This new empirical approach seems sufficiently good for use during the early screening; it gives reasonable estimates of CL(H) and good ranking, which allows replacement of C(bl)/C(pl) measurements by a simple equation.

Prediction of human pharmacokinetics--evaluation of methods for prediction of volume of distribution.

The aim was to evaluate and review methods for prediction of the steady-state volume of distribution (V(D,ss)) of xenobiotics in man. For allometry, approximately 30-40% of predictions are classified as incorrect, humans and animals belong to different V(D,ss) categories for approximately 30% of the compounds, maximum prediction errors are large (>10-fold), the b-exponent ranges between -0.2 and 2.2 (averaging approximately 0.8-0.9), and >2-fold prediction errors are found for 35% of the substances. The performance is consistent with species differences of binding in and outside the vasculature. The largest errors could potentially lead to very poor prediction of exposure profile and failure in clinical studies. A re-evaluation of allometric scaling of unbound tissue volume of distribution demonstrates that this method is less accurate (27% of predictions >2-fold errors) than a previous evaluation demonstrated. By adding molecular descriptor information, predictions based on animal V(D,ss) data can be improved. Improved predictions (approximately 1/10 of allometric errors) can also be obtained by using the relationship between unbound fraction in plasma (f(u,pl)) and V(D,ss) for each substance (method suggested by the author). A physiologically-based 4-compartment model (plasma, red blood cells, interstitial fluid and cell volume) together with measured tissue-plasma partitioning coefficients in rats, f(u,pl), interstitial-plasma concentration ratio of albumin, organ weight and blood flow data has been successfully applied. Prediction errors for one basic and one neutral drug are only 3-5%. The data obtained with this comparably laboratory-intensive method are limited to these two compounds. A similar approach where predicted tissue partitioning is used, and a computational model, give prediction errors similar to that of allometry. Advantages with these are the suitability for screening and avoidance of animal experiments. The evaluated methods do not account for potential active transport and slow dissociation rates.

Evaluation and suggested improvements of the Biopharmaceutics Classification System (BCS).

This review has evaluated the Biopharmaceutics Classification System (BCS) and improvements have been proposed. The BCS has a very strict solubility/dissolution limit, a generous P(e)-limit (> or = 14-times higher rate constant limit for dissolution than for permeation), and is stricter for drugs with a long half-life (t(1/2)). Available human in-vivo, in-vitro, and in-silico P(e)-methods cannot classify P(e) for moderately to highly permeable substances sufficiently well, and in-vitro data often underpredict the in-vivo dissolution potential and rate. Good in-vivo dissolution and absorption can be expected for most high P(e) drug products. It has not been possible to find a highly permeable product with a Dose number (D(o)) < 385 (< 2400 in the fed state) that is clearly incompletely absorbed, and near complete uptake has been shown for a drug product with a D(o) of 660000. The potential implication of these findings is that many true BCS Class I drug products are incorrectly classified. This could be a reason for the limited use of this system. On this basis, it has been suggested that: the limit for high for solubility/dissolution is decreased (to > 40 and > 95% dissolved within 30 min and 3 h, respectively); the limit for high P(e) is increased (to >P(e) of metoprolol); accurate P(e)-models or in-vivo fraction absorbed data are used; solubility/dissolution tests are performed using real or validated simulated gastrointestinal fluids; in-vitro/in-vivo dissolution relationships are established; the t(1/2) is considered; and the rate-limiting step for in-vivo absorption is determined. A major change could be to reduce the BCS into two classes: permeation-rate (Class I) or dissolution-rate (Class II) limited absorption. It is believed that this could give a better balance and increase the number of biowaivers.

Prediction of human pharmacokinetics--evaluation of methods for prediction of hepatic metabolic clearance.

Methods for prediction of hepatic clearance (CL(H)) in man have been evaluated. A physiologically-based in-vitro to in-vivo (PB-IVIV) method with human unbound fraction in blood (f(u, bl)) and hepatocyte intrinsic clearance (CL(int))-data has a good rationale and appears to give the best predictions (maximum approximately 2-fold errors; < 25% errors for half of CL-predictions; appropriate ranking). Inclusion of an empirical scaling factor is, however, needed, and reasons include the use of cryo-preserved hepatocytes with low activity, and inappropriate CL(int)- and f(u, bl)-estimation methods. Thus, an improvement of this methodology is possible and required. Neglect of f(u, bl) or incorporation of incubation binding does not seem appropriate. When microsome CL(int)-data are used with this approach, the CL(H) is underpredicted by 5- to 9-fold on average, and a 106-fold underprediction (attrition potential) has been observed. The poor performance could probably be related to permeation, binding and low metabolic activity. Inclusion of scaling factors and neglect of f(u, bl) for basic and neutral compounds improve microsome predictions. The performance is, however, still not satisfactory. Allometry incorrectly assumes that the determinants for CL(H) relate to body weight and overpredicts human liver blood flow rate. Consequently, allometric methods have poor predictability. Simple allometry has an average overprediction potential, > 2-fold errors for approximately 1/3 of predictions, and 140-fold underprediction to 5800-fold overprediction (potential safety risk) range. In-silico methodologies are available, but these need further development. Acceptable prediction errors for compounds with low and high CL(H) should be approximately 50 and approximately 10%, respectively. In conclusion, it is recommended that PB-IVIV with human hepatocyte CL(int) and f(u, bl) is applied and improved, limits for acceptable errors are decreased, and that animal CL(H)-studies and allometry are avoided.