A novel in vitro allometric scaling methodology for aldehyde oxidase substrates to enable selection of appropriate species for traditional allometry

1.?Failure to predict human pharmacokinetics of aldehyde oxidase (AO) substrates using traditional allometry has been attributed to species differences in AO metabolism.

2.?To identify appropriate species for predicting human in vivo clearance by single-species scaling (SSS) or multispecies allometry (MA), we scaled in vitro intrinsic clearance (CL_int) of five AO substrates obtained from hepatic S9 of mouse, rat, guinea pig, monkey and minipig to human in vitro CL_int.

3.?When predicting human in vitro CL_int, average absolute fold-error was ≤2.0 by SSS with monkey, minipig and guinea pig (rat/mouse >3.0) and was <3.0 by most MA species combinations (including rat/mouse combinations).

4.?Interspecies variables, including fraction metabolized by AO (F_m,AO) and hepatic extraction ratios (E) were estimated in vitro. SSS prediction fold-errors correlated with the animal:human ratio of E (r²?=?0.6488), but not F_m,AO (r²?=?0.0051).

5.?Using plasma clearance (CL_p) from the literature, SSS with monkey was superior to rat or mouse at predicting human CL_p of BIBX1382 and zoniporide, consistent with in vitro SSS assessments.

6.?Evaluation of in vitro allometry, F_m,AO and E may prove useful to guide selection of suitable species for traditional allometry and prediction of human pharmacokinetics of AO substrates.     

Reliability of human cryopreserved hepatocytes and liver microsomes as in vitro systems to predict metabolic clearance

A total of 110 drugs, selected to cover a range of physicochemical and pharmacokinetic properties, were used to explore standard approaches to the prediction of in vivo metabolic clearance using drug-depletion profiles from human liver microsomes (HLMs) and cyropreserved hepatocytes. A total of 41 drugs (37% of the compounds tested) showed measurable depletion rates using HLMs (depletion by 20% or more over the time course). The most reliable correlations in terms of bias (average fold error (AFE) = 2.32) and precision (root mean square error (RMSE) = 3501) were observed by comparing in vivo intrinsic clearance (CL_int), calculated using the parallel-tube model and incorporating the fraction unbound in blood, with in vitro CL_int adjusted for microsomal binding. For these reference drugs, 29% of predictions were within two-fold of the observed values and 66% were within five-fold. Compared with HLMs, clearance predictions with cryopreserved hepatocytes (57 drugs) were of similar precision (RMSE = 3608) but showed more bias (AFE = 5.21) with 18% of predictions within two-fold of the observed values and 46% within five-fold. However, with a broad complement of drug-metabolizing enzymes, hepatocytes catalysed measurable CL_int values for a greater proportion (52%) of the reference compounds and were particularly proficient at defining metabolic rates for drugs with predominantly phase 2 metabolic routes.     

Interspecies scaling of a camptothecin analogue: Human predictions for intravenous topotecan using animal data

As a class, camptothecin analogues via market entry of topotecan and irinotecan, have shown promise for the treatment of various solid tumours. Topotecan, in particular, was chosen as the substrate for allometric scaling and prediction of human parameter values for both total clearance (CL) and volume of distribution (V_ss). The availability of published data in mouse, rat, dog, and monkey paved the way for interspecies scaling via allometry. Although it appeared that at a minimum mouse, rat, and dog would reasonably fit in a three-species allometry scale-up, the inclusion of monkey data enabled a better prediction of the human parameter values for total topotecan–e.g., CL: allometric equation: 1.5234W^0.7865; predicted value = 43.04 l h^?1: observed CL = 24–53 l h^?1; V_ss: allometric equation: 1.1939W^1.0208; predicted value = 91.29 litres: observed V_ss = 66–146 litres. The proximity of the allometric exponent values of CL (0.7885) and V_ss (1.0208) to the suggested values of 0.75 and 1.00 was not only encouraging, but also confirmed the applicability of interspecies scaling approach for topotecan. The data suggest that allometric scaling approaches with suitable correction factors could potentially be used to predict the human pharmacokinetics of novel CPT analogues prospectively.     

Prediction of the pharmacokinetic parameters of reduced- dolasetron in man using in vitro-in vivo and interspecies allometric scaling

1. Dolasetron (Anzemet®) is apotent andselective 5-HT₃ receptor antagonist which is rapidly and extensively reduced to yield its major pharmacologically active metabolite, reduced dolasetron (RD). RD is further metabolized by CYP450 enzymes as well as undergoing renal excretion. As both in vitro and in vivo data on RD were available from animals andman, two approaches to predict the human pharmacokinetic parameters ofRD were assessed. 2. First, in vitro studies, using liver microsomes from animal species and man, were undertaken to measure V_max and K_m and to assess the intrinsic clearance (CL_int). With appropriate liver weight and liver bloodflow scaling factorsthe predicted in vivo metabolic clearance (CL_{m_pred}) was calculated. Human CL_{m_pred} was underestimated by a factor of 5 when it was calculated using the above scaling factors. As, in a prospective study, the observed human in vivo metabolic clearance (CL_{m_obs})is unknown, CL_{m_pred} was substituted into the least-squares correlation equation obtained from a plot of CL_{m_pred} against CL_{m_obs}, using animal data. The estimate of human CL_{m_obs} was improved as it was only underestimated by a factor of 1.5. 3. Second, allometric scaling of in vivo animal pharmacokinetic data, using body weight, was performed topredict pharmacokinetic parameters in man. Good predictions of human pharmacokinetic parameters of RD were obtained for plasma clearance (1.7 l/min predicted versus 1.6 l/min observed), half-life (6.0?h predicted versus 5.6?h observed), and volume of distribution (860.9 l predicted versus 770.4 l observed). 4. The integration of in vitro metabolic data from microsomes gave similar results to conventional allometric scaling, whereas the normalization of clearance by brain weight resulted in an approximately three-fold underestimation of human clearance. 5. For RD, a drug that is eliminated by both renal and metabolic clearance, retrospective conventional allometric scaling allowed accurate prediction of pharmacokinetic parameters in man, whereas in vitro-in vivo scaling resulted in an underestimation of in vivo CL_m. Although these results are somewhat at variance, ideally both scaling methods should be applied to improve the prediction of human pharmacokinetic parameters.     

Use of uptake intrinsic clearance from attached rat hepatocytes to predict hepatic clearance for poorly permeable compounds

1. We previously reported that the accuracy of clearance (CL) prediction could be differentiated by permeability. CL was drastically under-predicted by in vitro metabolic intrinsic clearance (CL_int) for compounds with low permeability (<5?×?10^?6 cm/s).

2. We determined apparent uptake CL_int by measuring initial disappearance from medium using attached rat hepatocytes and metabolic CL_int by measuring parent depletion in suspended rat hepatocytes (cells and medium).

3. Uptake and metabolic CL_int were comparable for highly permeable metabolic marker compounds. In contrast, uptake CL_int was 3- to 40-fold higher than metabolic CL_int for rosuvastatin, bosentan, and 15 proprietary compounds, which had low permeability, suggesting that uptake could be a rate-determining step in hepatic elimination for these poorly permeable compounds.

4. The prediction of hepatic CL was improved significantly when using uptake CL_int for the compounds with low permeability. The average fold error was 2.2 and 6, as opposed to >11 and >47 by metabolic CL_int, with and without applying a scaling factor of 4, respectively.

5. Uptake CL_int from attached hepatocytes can be used as an alternative approach to predict hepatic clearance and to understand the significance of hepatic uptake in elimination in an early drug discovery setting.

     

Prediction of human clearance of therapeutic proteins: simple allometric scaling method revisited

In this report, the utility of a commonly used interspecies scaling method to predict the systemic clearance (CL) of therapeutic proteins in humans was evaluated. Based on analysis of a pharmacokinetic data set of 34 therapeutic proteins, including 12 monoclonal antibodies (mAbs) and Fc fusion proteins, human CL can generally be predicted reasonably well with simple allometric scaling and a fixed exponent of 0.8:～95% of the cases predicted values within 2‐fold of the observed values when using CL data from multiple species, or～90% simply using CL from monkeys. Specific to mAbs/Fc fusion proteins, scaling from monkey CL using a fixed exponent of 0.8 gave an excellent prediction; all predicted CL values were within 2‐fold of the corresponding observed values. Compared with the simple allometric scaling method that uses a fitted exponent from CL data of ≥3 preclinical species, the fixed exponent approach with 1–2 preclinical species is simple, resource‐saving and minimizes systematic bias. Together with its overall satisfactory prediction accuracy, especially in the absence of non‐linear pharmacokinetics and species‐specific clearance mechanisms, this fixed exponent method affords a viable alternative to other published allometric methods, including the Rule of Exponents (ROE). Copyright © 2010 John Wiley & Sons, Ltd.     

Metabolite kinetics of ondansetron in rat. Comparison of hepatic microsomes,isolated hepatocytes and liver slices,with in vivo disposition

1. The kinetics of hydroxylation and N-demethylation of ondansetron have been determined in freshly isolated hepatocytes, hepatic microsomes and precision-cut liver slices from the male Sprague-Dawley rat. In vivo studies have also been carried out to characterize the pharmacokinetics of ondansetron and in vitro data have been assessed for their value as predictors of hepatic clearance.

2. In the three in vitro systems, the formation of hydroxylated and demethylated metabolites were characterized as a function of substrate concentration by a high-affinity, low-capacity site and a low-affinity, high-capacity site which was not saturated over the concentration range studied (2.5–500 μM). Slices gave consistently higher K_m's (20 and 30 μM for hydroxylation and demethylation respectively) than hepatocytes (3 and 13 μM respectively) and microsomes (2 and 5 μM respectively). The rank order of V_max and CL_int was the same for each system; hydroxylation rates exceeding demethylation rates. Although two hydroxylations (7- and 8-hydroxy metabolites) occurred exclusively in microsomes, these are believed to originate from a common precursor.

3. The high CL_int of ondansetron (150 ml/min/SRW, where SRW is a standard rat weight of 250 g) is well predicted by scaling either microsomal clearance for microsomal protein recovery or hepatocyte clearance for hepatocellularity (212 and 135 ml/min/SRW respectively). In contrast, the use of liver slice data scaled to a whole liver substantially underestimates CL_int (9 ml/min/SRW).     

Verapamil hepatic clearance in four preclinical rat models: towards activity‐based scaling

The current study was designed to cross‐validate rat liver microsomes (RLM), suspended rat hepatocytes (SRH) and the isolated perfused rat liver (IPRL) model against in vivo pharmacokinetic data, using verapamil as a model drug. Michaelis‐Menten constants (K_m), for the metabolic disappearance kinetics of verapamil in RLM and SRH (freshly isolated and cryopreserved), were determined and corrected for non‐specific binding. The ‘unbound’ K_m determined with RLM (2.8 µ m ) was divided by the ‘unbound’ K_m determined with fresh and cryopreserved SRH (3.9 µ m and 2.1 µ m , respectively) to calculate the ratio of intracellular to extracellular unbound concentration (Kp_u,u). Kp_u,u was significantly different between freshly isolated (0.71) and cryopreserved (1.31) SRH, but intracellular capacity for verapamil metabolism was maintained after cryopreservation (200 vs. 191 µl/min/million cells). Direct comparison of intrinsic clearance values (Cl_int) in RLM versus SRH, yielded an activity‐based scaling factor (SF) of 0.28–0.30 mg microsomal protein/million cells (MPPMC). Merging the IPRL‐derived Cl_int with the MPPMC and SRH data, resulted in scaling factors for MPPGL (80 and 43 mg microsomal protein/g liver) and HPGL (269 and 153 million cells/g liver), respectively. Likewise, the hepatic blood flow (61 ml/min/kg b.wt) was calculated using IPRL Cl_int and the in vivo Cl. The scaling factors determined here are consistent with previously reported CYP450‐content based scaling factors. Overall, the results show that integrated interpretation of data obtained with multiple preclinical tools (i.e. RLM, SRH, IPRL) can contribute to more reliable estimates for scaling factors and ultimately to improved in vivo clearance predictions based on in vitro experimentation. Copyright © 2015 John Wiley & Sons, Ltd.     

Approach to the prediction of the contribution of major cytochrome P450 enzymes to drug metabolism in the early drug-discovery stage

It is important to determine the cytochrome P450 (CYP) contribution of certain drugs by taking into consideration the attrition due to issues such as genetic polymorphism and inter-individual variation. In many cases in the early discovery stage, the metabolites of a new chemical have not been identified. Therefore, the present paper devised an approach in which the in vitro intrinsic clearance (CL_int) value for new chemicals was determined by measuring substrate depletion. The following prediction methods were compared to calculate CL_int using data from recombinant CYP enzymes: (1) the relative CYP content in human liver microsomes; (2) the relative activity factor (RAF) based on the V_max value; and (3) the RAF value based on the CL_int value. The most accurate prediction method was RAF based on CL_int. This method would be useful in the early drug-discovery process in cases in which the main metabolite is not identified.     

Comparison of predicted intrinsic hepatic clearance of 30 pharmaceuticals in canine and feline liver microsomes

1.?Known cytochrome P450 (CYP) substrates in humans are used in veterinary medicine, with limited knowledge of the similarity or variation in CYP metabolism. Comparison of canine and feline CYP metabolism via liver microsomes report that human CYP probes and inhibitors demonstrate differing rates of intrinsic clearance (CL_int).

2.?The purpose of this study was to utilize a high-throughput liver microsome substrate depletion assay, combined with microsomal and plasma protein binding to compare the predicted hepatic clearance (CL_hep) of thirty therapeutic agents used off-label in canines and felines, using both the well-stirred and parallel tube models.

3.?In canine liver microsomes, 3/30 substrates did not have quantifiable CL_int, while midazolam and amitriptyline CL_int was too rapid for accurate determination. A CL_hep was calculated for 29/30 substrates in feline microsomes. Overall, canine CL_hep was faster compared to the feline, with fold differences ranging from 2–20-fold.

4.?A comparison between the well-stirred and parallel tube model indicates that the parallel tube model reports a slighter higher CL_hep in both species.

5.?The differences in CYP metabolism between canine and feline highlight the need for additional research into CYP expression and specificity.     

Relative Importance of Intestinal and Hepatic Glucuronidation—Impact on the Prediction of Drug Clearance

Purpose  To assess the extent of intestinal and hepatic glucuronidation in vitro and resulting implications on glucuronidation clearance prediction. Methods  Alamethicin activated human intestinal (HIM) and hepatic (HLM) microsomes were used to obtain intrinsic glucuronidation clearance (CL_int,UGT) for nine drugs using substrate depletion. The in vitro extent of glucuronidation (fm_UGT) was determined using P450 and UGT cofactors. Utility of hepatic CL_int for the prediction of in vivo clearance was assessed. Results  fm_UGT (8–100%) was comparable between HLM and HIM with the exception of troglitazone, where a nine-fold difference was observed (8% and 74%, respectively). Scaled intestinal CL_int,UGT (per g tissue) was six- and nine-fold higher than hepatic for raloxifene and troglitazone, respectively, and comparable to hepatic for naloxone. The remaining drugs had a higher hepatic than intestinal CL_int,UGT (average five-fold). For all drugs with P450 clearance, hepatic CL_int,CYP was higher than intestinal (average 15-fold). Hepatic CL_int,UGT predicted on average 22% of observed in vivo CL_int; with the exception of raloxifene and troglitazone, where the prediction was only 3%. Conclusion  Intestinal glucuronidation should be incorporated into clearance prediction, especially for compounds metabolised by intestine specific UGTs. Alamethicin activated microsomes are useful for the assessment of intestinal glucuronidation and fm_UGT in vitro.     

Approach to predict the contribution of cytochrome P450 enzymes to drug metabolism in the early drug-discovery stage: The effect of the expression of cytochrome b5 with recombinant P450 enzymes

In order to evaluate the potential adverse effects due to genetic polymorphism and/or inter-individual variation, it is necessary to calculate the cytochrome P450 (CYP) contribution to the metabolism of new drugs. In the current study, the in vitro intrinsic clearance (CL_int) values of marker substrates and drugs were determined by measuring metabolite formation and substrate depletion, respectively. Recombinant CYP microsomes expressing CYP2C9, CYP2C19 and CYP3A4 with co-expressed cytochrome b₅ were used, but those expressing CYP1A2 and CYP2D6 did not have co-expressed cytochrome b₅. The following prediction methods were compared to determine the CL_int value using data from recombinant CYP enzymes: (1) relative CYP enzyme content in human liver microsomes; (2) relative activity factor (RAF) estimated from the V_max value; and (3) RAF estimated from the CL_int value. Estimating RAF from CL_int proved the most accurate prediction method among the three tested, and differences in the CYP3A4 marker reactions did not affect its accuracy. The substrate depletion method will be useful in the early drug-discovery stage when the main metabolite and/or metabolic pathway has not been identified. In addition, recombinant CYP microsomes co-expressed with cytochrome b₅ might be suitable for the prediction of the CL_int value.     

A new approach to predicting human hepatic clearance of CYP3A4 substrates using monkey pharmacokinetic data

1. Focusing on the genetic similarity of CYP3A subfamily enzymes (CYP3A4 and CYP3A5) between monkeys and humans, we have attempted to provide a single-species approach to predicting human hepatic clearance (CL_h) of CYP3A4 substrates using pharmacokinetic parameters in cynomolgus monkeys following intravenous administrations.

2. Hepatic intrinsic clearance (CL_int,h) of six CYP3A4 substrates (alprazolam, clonazepam, diltiazem, midazolam, nifedipine, and quinidine), covering a wide range of clearance, in monkeys correlated well with that cited in literature for humans (R = 0.90) with a simple equation of Y = 0.165X (Y: human CL_int,h, X: monkey CL_int,h, represented in mL/min/kg).

3. To verify the predictability of human CL_int,h, monkey CL_int,h of a test set of CYP3A4 substrates cited in literature (dexamethasone, nifedipine, midazolam, quinidine, tacrolimus, and verapamil) was applied to the equation and human CL_int,h was calculated. The human CL_int,h of all the substrates was predicted within 3-fold error (fold error: 0.35–2.77).

4. The predictability of human CL_h by our method was superior to common in vivo prediction methods (allometry and liver blood flow method). These results suggest that human hepatic clearance of CYP3A4 substrates can be predicted by applying cynomolgus monkey CL_int,h obtained following intravenous administrations in each laboratory to the simple equation.     

In vitro-in vivo correlation for intrinsic clearance for CP-409,092 and sumatriptan: a case study to predict the in vivo clearance for compounds metabolized by monoamine oxidase

1. Oxidative deamination of the GABA_A partial agonist CP-409,092 and sumatriptan represents a major metabolic pathway and seems to play an important role for the clearance of these two compounds.

2. Similar to sumatriptan, human mitochondrial incubations with deprenyl and clorgyline, probe inhibitors of monoamine oxidase B and monoamine oxidase A (MAO-B and MAO-A), respectively, showed that CP-409,092 was metabolized to a large extent by the enzyme MAO-A.

3. The metabolism of CP-409,092 and sumatriptan was therefore studied in human liver mitochondria and in vitro intrinsic clearance (CL_int) values were determined and compared to the corresponding in vivo oral clearance (CL_PO) values. The overall objective was to determine whether an in vitro-in vivo correlation (IVIVC) could be described for compounds cleared by MAO-A.

4. The intrinsic clearance, CL_int, of CP-409,092 was approximately 4-fold greater than that of sumatriptan (CL_int, values were calculated as 0.008 and 0.002?ml/mg/min for CP-409,092 and sumatriptan, respectively). A similar correlation was observed from the in vivo metabolic data where the unbound oral clearance, CL(u)_PO, values in humans were calculated as 724 and 178?ml/min/kg for CP-409,092 and sumatriptan, respectively.

5. The present work demonstrates that it is possible to predict in vivo metabolic clearance from in vitro metabolic data for drugs metabolized by the enzyme monoamine oxidase.

     

Mechanistic investigations into the species differences in pinometostat clearance: impact of binding to alpha-1-acid glycoprotein and permeability-limited hepatic uptake

1.?The plasma clearance of the first-in-class DOT1L inhibitor, EPZ-5676 (pinometostat), was shown to be markedly lower in human compared to the preclinical species, mouse, rat and dog.

2.?This led to vertical allometry where various interspecies scaling methods were applied to the data, with fold-errors between 4 and 13. We had previously reported the elimination and metabolic pathways of EPZ-5676 were similar across species. Therefore, the aim of this work was to explore the mechanistic basis for the species difference in clearance for EPZ-5676, focusing on other aspects of disposition.

3.?The protein binding of EPZ-5676 in human plasma demonstrated a non-linear relationship suggesting saturable binding at physiologically relevant concentrations. Saturation of protein binding was not observed in plasma from preclinical species. Kinetic determinations using purified serum albumin and alpha-1-acid glycoprotein (AAG) confirmed that EPZ-5676 is a high affinity ligand for AAG with a dissociation constant (K_d) of 0.24?μM.

4.?Permeability limited uptake was also considered since hepatocyte CL_int was much lower in human relative to preclinical species. Passive unbound CL_int for EPZ-5676 was estimated using a correlation analysis of logD and data previously reported on seven drugs in sandwich cultured human hepatocytes.

5.?Incorporation of AAG binding and permeability limited hepatic uptake into the well-stirred liver model gave rise to a predicted clearance for EPZ-5676 within 2-fold of the observed value of 1.4?mL min^?1?kg^?1. This analysis suggests that the marked species difference in EPZ-5676 clearance is driven by high affinity binding to human AAG as well as species-specific hepatic uptake invoking the role of transporters.     

Comparison of intrinsic metabolic clearance in fresh and cryopreserved human hepatocytes

1. We compared the intrinsic clearance (CL_int) of a number of substrates in suspensions of fresh and cryopreserved human hepatocytes from seven donors.

2. CL_int values for a cocktail incubation of phenacetin, diclofenac, diazepam, bufuralol, midazolam, and hydroxycoumarin were 4.9?±?3.4, 18?±?7.2, 5.1?±?4.9, 6.3?±?3.3, 9.8?±?5.8 and 22?±?14?μl min^?1/10⁶ cells, respectively, and they correlated well with corresponding CL_int values using cryopreserved hepatocytes from 25 different donors.

3. CL_int values of each cocktail substrate and 20 AstraZeneca new chemical entities were compared in fresh and cryopreserved hepatocytes from the same three donors. There was a statistically significant correlation between CL_int in fresh and cryopreserved hepatocytes for each of the three livers (p?int values was 1.03.

4. In conclusion, the results add further support to the use of cryopreserved human hepatocytes as a screening model for the intrinsic clearance of new chemical entities.

     

Prediction of Human Metabolic Clearance from <Emphasis Type="Italic">In Vitro</Emphasis> Systems: Retrospective Analysis and Prospective View

Purpose  

To provide a definitive assessment of prediction of in vivo CL _int from human liver in vitro systems for assessment of typical underprediction.     

Prediction of clinical pharmacokinetic parameters of linezolid using animal data by allometric scaling: applicability for the development of novel oxazolidinones

1.?Allometric scaling has previously been used as an effective tool for the prediction of human pharmacokinetic data. The pharmacokinetic data for linezolid, a novel oxazolidinone to treat Gram-positive pathogens, in mice, rats and dogs were subjected to simple allometric scaling. Generated allometric equations for parameters such as clearance (CL), volume of distribution (V_ss) and elimination rate constant (K₁₀) were used to predict human pharmacokinetic parameters including elimination half-lives. In addition, the human plasma concentration–time curve was simulated using a one-compartmental model.

2.?Application of simple allometry (Y?=?aW^b) for animal parameters such as CL, V_ss, and K₁₀ showed excellent allometric fit (r?≥?0.98). The allometric equations for CL, V_ss, and K₁₀ were??0.5465W^0.6595,??0.1369W^0.9246, and??0.4117W^–0.3139, respectively. The confidence in predictability of CL and V_ss parameters was particularly high since the allometric exponents of CL and V_ss almost approached the suggested values of 0.75 and 1.00, respectively.

3.?Animal pharmacokinetic parameters generated in the present authors’?laboratories for linezolid were in close agreement with reported literature values. The predicted human values for CL (4.68?l?h^?1), V_ss (37.07 litres), and K₁₀ (0.10?h^?1) were within the range observed for linezolid in the literature (CL?=?4?10.5 l?h^?1; V_ss?=?21???53 litres; K₁₀?=?0.09???0.3?h^?1). The human half-life (t_1/2) predicted using allometry (6.9?h) was similar to reported values in humans of 5?h. In summary, the retrospective analysis for linezolid suggests that allometric scaling can be used as a prospective tool for predicting human pharmacokinetic parameters of novel oxazolidinones.     

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 (CL_H) and intrinsic CL_H (CL_int) vary widely among studies. Reasons are not fully investigated and understood. The possibility to select favorable reference data for in vivo CL_H and CL_int and unbound fraction in plasma (f_u) is among possible explanations. The main objective was to investigate how reference data selection influences log in vitro and in vivo CL_int-correlations (r²). Another aim was to make a head-to-head comparison vs an in silico prediction method. Human hepatocyte CL_int-data for 15 compounds from two studies were selected. These were correlated to in vivo CL_int estimated using different reported CL_H- and f_u-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 CL_int-predictions.