Although the United States Food and Drug Administration has disclosed a list of drugs with pharmacogenomic biomarkers for drug labeling, there is limited information regarding pharmacogenomic-associated drugs in Japan. Such associations include genetic variants of uridine diphosphate glucuronosyltransferase 1A1 for irinotecan, nudix hydrolase 15 for thiopurine drugs, and cytochrome P450 (P450) 2C9 for siponimod. The effects of such genetic variants on drug concentrations are similar to those from drug interactions. Because of race and dosage differences, the relevance of pharmacogenomic associations in Asian populations requires confirmation. This white paper proposes that in vitro pharmacogenomic information can be used to predict human pharmacokinetics and to describe in drug labels the changes in blood concentrations by genetic variants. For P450 variants CYP2C9∗3, CYP2C19∗2, CYP2C19∗3, CYP2D6∗10, and CYP3A4∗16, we propose using the enzymatic activity parameters obtained from in vitro functional analysis of the drug-metabolizing enzymes for multiple substrate drugs to predict the effects of these variants on human pharmacokinetics. Consequently, in patients prescribed only a single drug, anything more than a "moderate effect" on plasma exposure should be mentioned as a caution in the drug labels; such effects are likely caused by enzyme polymorphisms resulting in similar effects to drug-drug interactions.

Subcutaneous administration is widely used as a clinical administration route for Fc-fusion proteins. However, predicting bioavailability (BA) in humans after subcutaneous administration is challenging due to multiple factors involved in the absorption process. This study aimed to elucidate the impact of degradation on the BA of Fc-fusion proteins. We established two measurement methods for each Fc-fusion protein: the Fc/Fc method, which recognizes the Fc region; and the Protein/Fc method, which recognizes both protein and Fc region. BA of dulaglutide and romiplostim in rats were 80.1 % and 99.4 % by the Fc/Fc method, and 35.0 % and 55.5 % by the Protein/Fc method, respectively. The lower BA with the Protein/Fc method indicates that the protein region undergoes degradation during the absorption process. In stability studies with rat skin homogenates and lymphatic fluid, degradation of the protein region for both dulaglutide and romiplostim was confirmed, which was inhibited by protease inhibitors. In contrast, abatacept and etanercept were stable in skin homogenates and lymphatic fluid, and their BA in rats were comparable between the Fc/Fc and Protein/Fc methods. This study indicates that the presence or absence of protease-mediated degradation during the absorption process is one of the factors affecting the BA of Fc-fusion proteins.

Streptomyces griseolus CYP105A1 showed hydroxylation activity at 25- and 1α-positions of vitamin D to produce an active form of vitamin D. We have succeeded in dramatically increasing its activity by site-directed mutagenesis. We also found that CYP105A1 and its mutants metabolize various non-steroidal anti-inflammatory drugs. In this study, we attempted to metabolize simvastatin (SV), an HMG-CoA reductase inhibitor, by CYP105A1 mutants, and compared their metabolism with that by human CYP3A4. The SV was converted into multiple metabolites by the CYP105A1 variants CYP105A1-R84A and CYP105A1-R84A/M239A, with the latter exhibiting significantly higher activity than the former. The metabolites were estimated to be 6'-hydroxy-SV, 3'-hydroxy-SV, 3″-hydroxy SV, 3',5'-dihydrodiol SV, and 6'-exomethylene SV. In addition, 6'-hydroxy-SV was non-enzymatically converted to 3'-hydroxy-SV under acidic conditions. The X-ray crystal structure of SV-bound CYP105A1-R84A suggested the formation of a 3',4'-epoxide intermediate, from which 3'-hydroxy and 3',5'-dihydrodiol SV were presumed to be generated non-enzymatically. It is noted that all of these metabolites have either been reported as SV metabolites formed by human CYP3A4 or were detected in the present study. Thus, CYP105A1-R84A/M239A appears to be highly useful for the production of human metabolites of SV.

Ezetimibe is a widely used antilipemic agent that lowers LDL-C and raises HDL-C by inhibiting intestinal cholesterol absorption. However, the mechanisms governing the hepatic transport of ezetimibe and its glucuronide metabolite remain unclear. In this study, we demonstrated that ezetimibe is rapidly metabolized to its glucuronide in mouse, rat, and human liver microsomes, with UGT1A1 showing the highest catalytic activity, followed by UGT1A3. In vitro transport assays revealed that human OATP1B1 and OATP1B3 selectively mediate the uptake of ezetimibe-glucuronide, but not parent ezetimibe. Using Slco1b2 knockout mice, an in vivo model for hepatic OATP deficiency, we found that knockout significantly increased the plasma AUC and decreased the liver-to-plasma ratio of ezetimibe-glucuronide, without affecting parent ezetimibe. These results indicate that OATP-mediated hepatic uptake is key to the disposition of ezetimibe-glucuronide, which may influence the pharmacokinetics and clinical efficacy of ezetimibe.

Drug-induced liver injury often arises from reactive metabolites (RMs) produced in the liver, making it crucial to assess RM formation rates from drug candidates. Conventional assays using glutathione (GSH) effectively trap soft electrophilic RMs but fail to detect hard electrophiles. To address this, we developed a double trapping assay employing [H]GSH and [C]cyanide as soft and hard nucleophilic reagents, respectively. This assay was applied to 25 drugs chosen based on safety profiles. Eight drugs were exclusively trapped by [H]GSH, while 11 were trapped by [C]cyanide or both reagents, demonstrating that a double trapping assay provides a more comprehensive detection method for both soft and hard RMs. Multiplying RM formation rates by daily doses allowed almost complete differentiation between withdrawn/black boxed warning drugs and safer ones. Radio-LCMS analysis provided detailed insights into the substructures of drug candidates responsible for RM production. Interestingly, it was discovered that GSH-based assays sometimes fail to detect certain RMs due to the presence of dithiothreitol in commercial [H]GSH. This study highlights the efficacy of the double trapping assay using [H]GSH and [C]cyanide in accurately and comprehensively detecting RMs. Furthermore, it offers valuable structural information to minimize RM formation during early drug discovery.

Peptide drugs are expected to be a new modality that will replace traditional small molecule drugs. As the number of approved peptide drugs increases, they are being co-administered with various drugs, but there is a limited number of reports on their drug-drug interaction (DDI) in both in vitro and in vivo (clinical) studies. In this study, we investigated the transporter inhibitory potential of Compound A, a macrocyclic peptide (3.5 kDa) for the treatment of pain. We found that Compound A exhibited a strong inhibitory effect on the organic anion transporting polypeptide (OATP) 1B in an in vitro study. To assess the in vivo OATP1B inhibitory potential, Compound A was intravenously or subcutaneously administered to monkeys, and the plasma concentration of coproporphyrin-I (CP-I), an endogenous biomarker of OATP1B, was determined. Compound A markedly increased the CP-I concentration in monkeys. A semi-mechanistic pharmacokinetic model analysis using the CP-I concentration revealed that Compound A is a highly potent in vivo OATP1B inhibitor (in vivo K: 59.9 ng/mL as total plasma concentration). Our findings suggest that even peptides with a large molecular weight can cause DDI. These results offer valuable information for the further development of DDI guidelines for peptides.

The previously reported Template system for the prediction of human CYP2D6-mediated reactions (Drug Metab Dispos 40 486-96, 2012) has been refined with the introduction of ideas of allowable width, Trigger∗-residue and the residue-initiated movement of ligands in the active site. These ideas are in common with published Template systems for human CYP1A1, CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C18, CYP2C19, CYP2E1, CYP2J2, and CYP3A4/5/7 (Drug Metab Pharmacokinet 2016, 2019, 2020, 2021, 2022, 2023, and 2024, Food Safety 2024). Total 616 reactions of 441 distinct chemicals reported as CYP2D6 ligands were examined in the refined CYP2D6-Template system. From their placements on the refined Template and rules for interaction modes, verifications of good and poor substrates, regio/stereo-selectivity, and inhibition became faithfully available for these ligands. A comparison of the placements suggested key interactions with Shelf and Left-end for ligand accommodations on the refined CYP2D6-Template. Shelf-surrounding of ligands was also proposed as a cause of their intense inhibitions. The refined CYP2D6-Template system will thus offer reliable estimations of this human CYP catalyses toward ligands of diverse structures, together with their deciphering information to lead to judgments of regioselective metabolisms.

Cytochrome P450 (P450 or CYP) 2J2, which metabolizes exogenous medicines and endogenous arachidonic acid to 14,15-epoxyeicosatrienoic acid (14,15-EET), is expressed in various organs and cancer cells. Additionally, CYP2C8 catalyzes the synthesis of 14,15-EET, a vasodilator that promotes cancer cell proliferation. However, the effect of tyrosine kinase inhibitors (TKIs) used in leukemia treatment on CYP2J2 and CYP2C8 remains unclear. This study investigated the effects of 16 TKIs used for leukemia treatment on recombinant CYP2J2-and CYP2C8-mediated processes. Among the TKIs, nilotinib and radotinib strongly inhibited CYP2J2-dependent astemizole O-demethylation and rivaroxaban hydroxylation, and CYP2C8-mediated paclitaxel 6α-hydroxylation (<20 %), with competitive inhibition constants of 0.41 and 0.22 μM, respectively (for astemizole O-demethylation). Nilotinib and radotinib suppressed CYP2J2-and CYP2C8-catalyzed arachidonic acid epoxidation and decreased their mRNA expression in Huh-7 cells (possibly via the peroxisome proliferator-activated receptor α pathway). Given that their inhibition constants are lower than their reported plasma concentrations, both may substantially suppress CYP2J2 and CYP2C8 functional enzyme levels and enzymatic activities in clinical settings. This suppression could potentially alter vasodilation by affecting 14,15-EET production, influencing CYP2J2 and CYP2C8-mediated drug-disease (conditions) and drug-drug interactions.

Gabapentin is an anticonvulsant used in the pharmacotherapy of pediatric epilepsy. Differences in transporter-mediated cerebral drug/compound distribution across the blood-brain barrier (BBB) have been reported between neonates/infants, and adults. The purpose of our study was to comprehensively elucidate the transport of gabapentin across the BBB during childhood and its regulatory molecular systems. The cerebral distribution of [H]gabapentin in 7-day-old rats was significantly lower than that in 42-day-old adult rats, suggesting weaker gabapentin transport in the BBB of neonates/infants than in adults. In vivo brain uptake and in vitro transport studies have indicated the involvement of transporters that accept large neutral amino acids in the cerebral gabapentin distribution across the BBB. In particular, an in vitro study using RNA interference suggested a major contribution of L-type amino acid transporter 1 (LAT1) to gabapentin transport across the BBB. In the brains of 7-day-old rats, the mRNA expression of LAT1 and the heavy chain of the 4F2 antigens (4F2hc), which forms a complex with LAT1, was reduced compared to that of 42-day-old rats. Consequently, it is suggested that a decrease in gabapentin distribution to the brain across the BBB contributes to the transcriptional reduction of cerebral LAT1 and 4F2hc.

Resistance to drugs used at the onset of follicular lymphoma could develop after continued drug treatment or at the time of recurrence. Therefore, we investigated how the continuous exposure to therapeutic drugs affects P-gp and drug resistance using follicular lymphoma cells. The mRNA expression levels of P-gp in Sci-1 cells treated with Dox or Dox + Vinc were increased. The increase in protein levels was consistent with the behavior of mRNA. The efflux clearance of Rho123 in Sci-1 cells exposed to Dox or Dox + Vinc was significantly increased. In cells continuously exposed to Dox or Dox + Vinc, a significant increase in the IC50 of Dox and an inhibitory effect of Vera on this increase were observed. It has been revealed that therapeutic drugs for follicular lymphoma, such as Dox or Dox + Vinc, increase the expression level of P-gp at the transcriptional level. The transport function of P-gp and drug resistance were enhanced depending on the increase in the amount of P-gp expressed on the membrane. Consequently, our results suggested that continuous exposure to clinical blood levels of Dox + Vinc can significantly affect drug resistance in follicular lymphoma by enhancing P-gp function through transcriptional and post-transcriptional regulation.

Drug-induced liver injury (DILI) presents a major challenge not only in new drug development but also in post-marketing withdrawals and the safety of food, cosmetics, and chemicals. Experimental model organisms such as the rodents have been widely used for preclinical toxicological testing. However, the tension exists associated with the ethical and sustainable use of animals in part because animals do not necessarily inform the human-specific ADME (adsorption, dynamics, metabolism and elimination) profiling. To establish alternative models in humans, in vitro hepatic tissue models have been proposed, ranging from primary hepatocytes, immortal hepatocytes, to the development of new cell resources such as stem cell-derived hepatocytes. Given the evolving number of novel alternative methods, understanding possible combinations of cell sources and culture methods will be crucial to develop the context-of-use assays. This review primarily focuses on 3D liver organoid models for conducting. We will review the relevant cell sources, bioengineering methods, selection of training compounds, and biomarkers towards the rationale design of in vitro toxicology testing.

ABCG2 (breast cancer resistance protein, BCRP), can affect drug disposition, and thus, variation in the ABCG2 gene may alter drug exposure. We studied non-synonymous naturally occurring single-nucleotide variants (SNVs) in intracellular loop 1 (ICL1), which contains a coupling helix that transmits conformational changes in the protein. Reference ABCG2, the common SNVs V12M and Q141K, and five SNVs (K453R, I456V, H457R, G462R and G462V) in ICL1 were expressed in HEK293 cells. Additionally, combinations of selected SNVs were expressed to determine if an activating substitution in ICL1 could compensate for an inactivating substitution elsewhere. Transport of Lucifer yellow, estrone sulfate and rosuvastatin was studied using membrane vesicles and the ABCG2 abundance was quantified. While K453R and I456V abundance was similar to the reference ABCG2, abundance was lower for H457R and G462R/V. Apparent transport activities were partially substrate dependent, but excluding G462R/V, the ICL variants transported at least one of the substrates similarly to the reference ABCG2. In double substitutions, I456V had a more consistent effect than H457R on both transport activity and protein abundance. Altogether, SNVs in ICL1 can have both detrimental and beneficial effects on ABCG2 activity. Effects may be hard to predict, especially if more than one SNV is present.

Imaging mass spectrometry (IMS) is used in various fields of pharmaceutical research and development, including the targeted delivery of administered drugs, drug distribution in tissues, drug toxicity analysis, and disease mechanisms. However, IMS is a relatively new technology that requires further validation before being accepted by authorities for regulatory compliance of new drugs. In 2020, an international survey was conducted in collaboration with the Imaging Mass Spectrometry Society (IMSS) and the Japan Association for Imaging Mass Spectrometry (JAIMS) to understand the status of IMS and to identify problems with its application. The survey revealed technical challenges in sample preparation, quantitative analysis of drugs in tissues, and data acquisition. Based on the 2020 survey results, we recently conducted a further detailed survey and had discussions within the JAIMS aimed at translating the results into specific experimental procedures and proposing feasible standard methods. This survey involved detailed questions on five themes: sample collection and storage, tissue section preparation, sample preparation, data analysis (including quantitative analysis and data correction methods), and data reproducibility. The questions were answered by JAIMS members working for ten companies. To resolve technical issues identified by the survey, we propose some realistic approaches toward standardization.

We systematically evaluated the effects of different buffer compositions on the activities of cytochrome P450 (CYP)1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, and CYP3A in human liver microsomes and in recombinant human CYP expression systems using the in vitro cocktail method. The activities of all CYP isoenzymes varied depending on the buffer conditions, particularly the activity of CYP1A2, which varied by 27.0-fold in phosphate buffer and HEPES or Tris-HCl buffer. The difference in the metabolic activities of CYP2D6 and CYP3A4 was smaller but significant, showing values about 2.0- and 3.0-fold higher in phosphate buffer than in HEPES or Tris-HCl buffer. A similar trend was observed for recombinant human CYP3A4/5. In contrast, the metabolic activities of CYP2A6 and CYP2B6 in phosphate buffer decreased by less than half as the phosphate concentration was increased. Changes in the metabolic activity of multiple CYP isoenzymes in buffers with different compositions suggest that the fraction metabolized (fm) evaluated in vitro differs depending on the experimental conditions. Specifically, the previously reported fm values of CYP1A2, CYP2B6, CYP2D6, and CYP3A in phosphate buffer may be overestimated. When evaluating the fm of these isoenzymes, we recommend conducting experiments under multiple buffer conditions.

Equilibrative nucleoside transporters (ENTs) are known to be involved in the membrane transport of nucleosides and nucleobases. We here report our finding that ENT2/SLC29A2 can also transport urate, which is the final metabolite of purine nucleobases. In transiently transfected human embryonic kidney 293 (HEK293) cells, urate uptake by human ENT2 was quite efficient at an acidic pH of 5.5, being saturable with the K of 1.64 mM and inhibited by its nucleoside substrates and specific inhibitors. Although the uptake activity decreased with increasing pH, it was still detected at around near neutral pH. In the Caco-2 cell model, urate uptake was reduced by ENT2 inhibitors and by ENT2 knockdown, indicating the involvement of ENT2 in urate transport. In HEK293 cells transiently transfected with sodium-dependent nucleobase transporter 1, which can accumulate urate, urate uptake was reduced by cotransfection of ENT2, suggesting its operation for urate efflux. Interestingly, ENT2 with N68K mutation was found to have little urate uptake activity, whereas the urate efflux activity was maintained. This finding suggests that ENT2 may operate in different modes for the uptake and efflux of urate. Overall, this study provides novel insight into the function of ENT2 and its potential role in urate disposition.

Preclinical animal models cannot precisely predict medication failures in clinical trials. Human primary hepatocytes (HPHs) are the gold standard for in vitro culture systems that recapitulate in vivo human liver characteristics. Consequently, demand for HPHs is rising despite their high cost, batch-to-batch variability, and inadequate supply. Existing commercial HPHs are plateable (pHPHs) and suspension (sHPHs) using same isolation and cryopreservation technique. The sHPHs cannot be cultured on the plate surface or cannot efficiently relocalize efflux transporters. They can also have a life of maximum of 2-4 h. Therefore, they are unsuitable for assays requiring well-organized efflux transporters or long-term studies, such as enzyme induction or metabolic assays for low-turnover drugs. We aim to develop a novel pharmacokinetic model of converted sHPHs to be attached like pHPHs. We successfully cultured sHPHs as 2D sandwich-cultured sHPHs for use as a promising, reproducible, and affordable in vitro model for efflux trans-porter assays, cytochrome P450 induction, and metabolic studies. The sHPHs successfully attached, converted to be plateable, similar morphology, and mRNA expression as pHPHs, as well as detectable metabolic activity and extended lifespan. Overall, this study is the first to culture functional 2D sandwich sHPHs as a novel in vitro pharmacokinetic model.

The aldo-keto reductase (AKR) 1C subfamily, comprising AKR1C1-1C4, plays a crucial role in drug metabolism and hormone biosynthesis. Recent studies have identified AKR1C3 as a co-activator of the androgen receptor. This study aimed to investigate whether AKR1Cs function as regulators of the pregnane X receptor (PXR), a member of the nuclear receptor superfamily, which upregulates drug-metabolizing enzymes such as cytochrome P450 (CYP) 3A4. Rifampicin-activated CYP3A4 induction was attenuated by AKR1C1/1C2 knockdown in ShP51 cells (PXR-overexpressing HepG2 cells), HepaRG cells, and HepaSH cells (hepatocytes from humanized liver mice). Co-immunoprecipitation analysis revealed that AKR1Cs interact with PXR. Immunofluorescent staining revealed that AKR1Cs are translocated into the nucleus with PXR by rifampicin in HepaRG cells. These results suggested that AKR1C1/1C2 has an ability to enhance transactivity of PXR. Consistent with the results of knockdown experiments, PXR-mediated CYP3A4 induction was significantly attenuated by treatment with AKR1C1/1C2 inhibitors, diazepam or flunitrazepam, in ShP51, HepaRG, and HepaSH cells. Furthermore, the induction of CYP2B6, CYP2C9, glucose 6-phosphatase, and phosphoenolpyruvate carboxykinase 1, all regulated by PXR, was attenuated by AKR1C1/1C2 inhibitors. Collectively, we demonstrated that AKR1C1/1C2 upregulates PXR transactivation. Clinically used drugs that inhibit AKR1C1/1C2 may suppress PXR-mediated transactivation of genes encoding drug-metabolizing and gluconeogenesis enzymes.

Chimeric mice with humanized liver (humanized-liver mice) are an attractive alternative to conventional animals for predicting pharmacokinetics in humans. We aimed to investigate the role of CYP2A6 and PON1 on pilocarpine biotransformation in humanized-liver mice. Following a single oral dose of 10 mg/kg, higher plasma concentrations of pilocarpine and 3-hydroxypilocarpine were observed in humanized-liver mice than in non-humanized mice. The ratios of area under the curve (AUC) to total AUC for pilocarpine in humanized-liver mice were more similar to those observed in humans with extensive metabolizer phenotypes for CYP2A6 compared to those in non-humanized mice for both 3-hydroxypilocarpine and pilocarpic acid. Pilocarpine 3-hydroxylase activity in liver microsomes was higher in humanized-liver mice than in non-humanized mice, whereas the pilocarpine hydrolase activity in plasma was comparable between both groups. CYP2A6 levels and pilocarpine 3-hydroxylase activities in liver microsomes from humanized-liver mice were correlated (P = 0.04). Furthermore, both hepatic microsomal 3-hydroxylase and plasma hydrolase activities of pilocarpine in humanized-liver mice decreased in the presence of respective human CYP2A6 or PON1 inhibitors. The plasma metabolite profiles of pilocarpine in humanized-liver mice were similar to those in humans, highlighting the suitability of humanized-liver mice for investigating CYP2A6-and PON1-dependent drug biotransformation in humans.

In vitro blood-brain barrier (BBB) models, primarily consisting of brain microvascular endothelial cells (BMEC), are expected to play pivotal roles in evaluating drug permeability into the brain. However, these models often exhibit functional variability due to various factors, raising practical concerns that can hinder their use in drug development studies. By investigating cell passage numbers as one such factor, we aim to assess how BBB model functionality is affected and to propose a practical strategy for managing this variability. In transwell-BBB models - but not in spheroidal-BBB models - the intercellular barrier integrity was somewhat compromised when higher-passage human immortalized BMEC (HBMEC/ci18) were used. Nonetheless, a clear in vitro-in vivo correlation (IVIVC) curve could still be obtained with these transwell-BBB models, similar to those with lower passage number HBMEC/ci18, presumably allowing for reasonable estimation of in vivo drug permeability. Therefore, changes in functional levels of BBB models do not always significantly diminish their practical value in drug BBB permeability studies. Additionally, the IVIVC curve integrity may serve as an indicator for assessing acceptable BBB model functionality. These findings provide valuable insights for the future application of in vitro human BBB models in drug development studies.

Brexpiprazole, widely approved for the treatment of schizophrenia, is an atypical antipsychotic that modulates serotonin-dopamine activity. To better understand the pharmacokinetics (PK) of brexpiprazole in Japanese patients, a population PK model was constructed and used to estimate steady state PK profiles and parameters as well as dopamine D/D receptor occupancy profiles after repeated oral administrations of brexpiprazole at 1 and 2 mg/day. Nonlinear mixed effects modelling was used to analyse data from a total of 398 healthy subjects and patients with schizophrenia who received brexpiprazole in three Japanese clinical trials. The PK of brexpiprazole were well described by a two-compartment disposition model with transit absorption compartments. Estimated glomerular filtration rate, age and cytochrome P450 2D6 phenotype were identified as significant covariates on CL/F only. The model predicted that, at a dose of 2 mg/day, trough plasma concentration (90 % prediction interval) of brexpiprazole is 77.1 (22.4-173) ng/mL and that dopamine D/D receptor occupancy is >80 % over one day for most patients at steady state. This suggests the recommended maintenance dose of 2 mg/day of brexpiprazole leads to clinically useful dopamine D/D receptor occupancy at steady state in Japanese patients.

We have experienced a 2.3-fold increase in the trough concentration of orally administered tacrolimus in a patient for 7 days after intravenous administration of trastuzumab emtansine (T-DM1), an antibody-drug conjugate. Retrospective clinical study revealed no change in factors known to alter the pharmacokinetics of tacrolimus, and the change in the trough concentration of tacrolimus was reversed after switching T-DM1 to trastuzumab (Tmab). Thus, interaction with T-DM1 was suspected as the cause of the increased trough concentration of tacrolimus. An animal study in rats showed that T-DM1 significantly increased the AUC of orally administered tacrolimus by more than 2-fold on day 0, day 3, and day 7, whereas no change was observed in the case of intravenous administration of tacrolimus. T-DM1 also significantly increased F・F of tacrolimus by more than 2-fold on day 7. In contrast, Tmab itself had no effect on the blood concentration of tacrolimus. These results suggest that T-DM1 increased the blood concentration of orally administered tacrolimus, and the effect persisted for 1 week after T-DM1 administration. Metabolites of T-DM1 excreted via the biliary route may contribute to the increase in the gastrointestinal absorption of tacrolimus.