Abstract
Aims
This phase 1, open‐label, crossover study sought to evaluate drug–drug interactions between tivantinib and cytochrome P450 (CYP) substrates and tivantinib and P‐glycoprotein.
Methods
The effect of tivantinib doses on the pharmacokinetics of the probe drugs for CYP1A2 (caffeine), CYP2C9 (S‐warfarin), CYP2C19 (omeprazole), and CYP3A4 (midazolam), and for P‐glycoprotein (digoxin) was investigated in 28 patients with advanced cancer using a cocktail probe approach. Patients received single doses of probe drugs alone and, after 5 days of treatment, with tivantinib 360 mg twice daily.
Results
The ratios of geometric least squares mean (90% confidence interval) for the area under the concentration–time curve from time zero to the last quantifiable concentration in the presence/absence of tivantinib were 0.97 (0.89–1.05) for caffeine, 0.88 (0.76–1.02) for S‐warfarin, 0.89 (0.60–1.31) for omeprazole, 0.83 (0.67–1.02) for midazolam, and 0.69 (0.51–0.94) for digoxin. Similar effects were observed for maximum plasma concentrations; the ratio for digoxin in the presence/absence of tivantinib was 0.75 (0.60–0.95).
Conclusions
The data suggest that tivantinib 360 mg twice daily has either a minimal or no effect on the pharmacokinetics of probe drugs for CYP1A2, CYP2C9, CYP2C19 and CYP3A4 substrates, and decreases the systemic exposure of P‐glycoprotein substrates when administered with tivantinib.
Keywords: cancer, cytochrome P450, drug interaction, P‐glycoprotein, pharmacokinetics, tivantinib
What is Already Known about this Subject
Tivantinib is a c‐MET receptor tyrosine kinase inhibitor currently under clinical investigation in patients with cancer.
Patients with cancer are usually on multiple therapies; many of the commonly used drugs are cytochrome P450 (CYP) and P‐glycoprotein (P‐gp) substrates.
The effect of tivantinib on the pharmacokinetics of CYP and P‐gp substrate in cancer patients is poorly understood.
What this Study Adds
Tivantinib 360 mg twice daily has either a minimal or no effect on the pharmacokinetics of CYP probe drugs, and decreases the systemic exposure of P‐gp substrates.
These results suggest that the drug interaction of tivantinib 360 mg twice daily is of minimal clinical significance when CYP1A2, CYP2C9, CYP2C19 and CYP3A4 substrates are concomitantly administered in cancer patients.
Close monitoring might be necessary, in terms of loss of efficacy, for P‐gp substrates with a narrow therapeutic window when concomitantly administered with tivantinib.
Introduction
Mutations in the receptor tyrosine kinase mesenchymal‐epithelial transcription factor (c‐MET) are a common feature of many carcinomas, and aberrant activation of c‐MET is involved in survival, invasion and metastasis of tumour cells 1. For this reason, c‐MET is an attractive target for anticancer drug development 1.
Tivantinib (ARQ 197) is an orally administered MET inhibitor that does not compete with adenosine triphosphate. It inhibits growth and induces apoptosis in MET‐expressing breast, colon, gastric and lung cancer cell lines and xenografts 1, 2. Several studies also demonstrate that tivantinib has antiproliferative activity in cell lines lacking c‐MET activity. These studies indicate that tivantinib impairs microtubule assembly 3, 4 and enhances apoptosis by inhibiting Mcl‐1 and Bcl‐xL and increasing cyclin B1 expression 5, 6. Tivantinib has been evaluated as monotherapy and in combination with other anticancer agents among patients with solid tumours 7. In particular, tivantinib is well tolerated and has shown antitumour activity in patients with high MET expression 8, 9. The safety and efficacy of tivantinib monotherapy on overall survival is currently being evaluated in a Phase 3 study in patients with hepatocellular carcinoma with high MET expression (ClinicalTrials.gov identifier: NCT01755767) 10.
The pharmacokinetics (PK) of tivantinib have been characterized in healthy subjects and in patients with cancer. After administration of a single 14C‐labelled oral dose of tivantinib to healthy subjects, most radioactivity was excreted via the biliary‐faecal route (68%), whereas 19% was excreted in urine 11. Tivantinib undergoes extensive metabolism. The majority of an administered dose was detected as metabolites beyond 8 h after oral administration. Seven metabolic pathways have been identified, including six that involve oxidation and one glucuronidation. Four metabolites have been shown to have some pharmacological activity, albeit considerably (3.3–8.4‐fold) less than the parent molecule 11.
When administered twice daily (BID) as an oral formulation in adult patients with solid tumours, exposure to tivantinib, as indicated by the area under the concentration–time curve (AUC), increases with increasing dose, although the relationship between dose and exposure is not dose proportional 12, 13. Indeed, at a dose of 400 mg BID, the AUC was nearly three‐fold greater than predicted for a dose‐proportional increase, suggesting that the drug elimination capacity was saturated at a dose of 400 mg BID 13. Considerable inter‐patient variability has been observed in adults and children treated with tivantinib over a wide dose range 12, 14, 15, 16. The elimination half‐life in adult patients with solid tumours ranges from 1.5 – 6.3 h 12, 13.
Cytochrome P450 (CYP) 2C19 is the key metabolic enzyme involved in elimination of tivantinib, and the PK vary with CYP2C19 phenotype 14, 15, 16. Drug exposure (AUC from time zero to 12 h [AUC 0–12]) increases in a dose‐dependent manner up to a dose of 300–360 mg BID in extensive metabolizers (EMs). In contrast, at a dose of 240 mg BID, exposure (AUC 0–12) in poor metabolizers (PMs) was approximately two‐fold that in EMs and similar to that achieved in EMs at a dose of 300–360 mg BID 17. Evaluation of the potential for drug–drug interactions (DDIs) is an important consideration in the development of any new drug 18. Many tyrosine kinase inhibitors (TKIs) alter the activity of cytochrome P450 (CYP) isozymes or modify P‐glycoprotein (P‐gp)‐mediated transmembrane transport and are involved in clinically significant DDIs 19, 20. An efficient approach to evaluate a candidate drug's potential for DDIs is the administration of a mixture of drugs in a “cocktail” 18, which allows for the simultaneous assessment of multiple drug–substrate interactions under controlled circumstances 21, 22. For example, use of the clinically validated Cooperstown 5 + 1 cocktail involves coadministration of five probe substrates, consisting of caffeine, S‐warfarin (plus vitamin K), omeprazole, dextromethorphan, and intravenous midazolam, to simultaneously assess the enzymatic activities for CYP1A2, CYP2C9, CYP2C19, CYP2D6, CYP3A, xanthine oxidase, and N‐acetyltransferase 2 in humans 22.
Preclinical studies in human hepatic microsomes showed that tivantinib has the potential to inhibit CYP activity. Competitive inhibition was observed for S‐mephenytoin 4′‐hydroxylation (CYP2C19, inhibitory constant [Ki] = 5.6 μM), bupropion hydroxylation (CYP2B6, Ki = 12.2 μM) and phenacetin O‐deethylation (CYP1A, Ki = 14.9 μM); noncompetitive inhibition was observed for diclofenac 4′‐hydroxylation (CYP2C9, Ki = 12.5 μM); and mixed noncompetitive inhibition was observed for amodiaquine N‐deethylation (CYP2C8, Ki = 8.7 μM), testosterone 6β‐hydroxylation (CYP3A, Ki = 20.5 μM) and midazolam 1′‐hydroxylation (CYP3A, Ki = 37.3 μM). In contrast, coumarin 7‐hydroxylation (CYP2A6), bufuralol 1′‐hydroxylation (CYP2D6), and chlorzoxazone 6‐hydroxylation (CYP2E1) did not meet the criterion for inhibition (half‐maximal inhibitory concentration [IC50] could not be calculated) (unpublished data on file, Daiichi Sankyo Co., Ltd., Tokyo, Japan). In a subsequent in vitro study in human hepatocytes, tivantinib 3 μM induced CYP3A4 mRNA transcription and enzyme activity (unpublished data on file, Daiichi Sankyo Co., Ltd.). Preliminary experiments using an in vitro system also show that two of the major metabolites of tivantinib (HPM6 and HPM8) have the potential to inhibit select CYP isoenzymes (unpublished data on file, Daiichi Sankyo Co., Ltd.). In a Caco‐2 cell monolayer, tivantinib showed an inhibitory effect on the efflux transporter P‐gp with an IC50 of 47.7 μM (unpublished data on file, Daiichi Sankyo Co., Ltd.).
On this basis, a dedicated DDI study using a cocktail approach was undertaken in patients with advanced cancer with the primary objective of determining the effect of multiple doses of tivantinib on the single‐dose PK of caffeine, S‐warfarin, omeprazole, midazolam, and digoxin when coadministered with tivantinib. As tivantinib had CYP3A4 inhibition and induction potential, a multiple‐dose study was needed to evaluate the effect of tivantinib on CYP3A4 substrates. Major metabolites of tivantinib have not been characterized for genotoxicity; therefore, after identification of the major metabolites, tivantinib was administered for study in patients with cancer.
Methods
This Phase 1 crossover study (ClinicalTrials.gov Identifier: NCT01517399) with an open‐label, single‐sequence design was conducted at three sites in the United States (South Texas Accelerated Research Therapeutics [START], San Antonio, TX; Duke University Medical Center, Durham, NC; and Vanderbilt University Medical Center, Nashville, TN). The study was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines. The protocol was approved by the local institutional review board at each centre (START: IRB00001035, IRB00003657, IRB00004920, IRB00006075; Duke University: Pro00032884; Vanderbilt University: 120168), and written informed consent was obtained from each patient before undergoing any study procedure.
Patients
Male or female adults aged ≥18 years with histologically or cytologically confirmed advanced solid tumours at screening were eligible for the study. In addition, eligible patients were required to have an Eastern Cooperative Oncology Group performance status ≤2 and adequate bone marrow (haemoglobin ≥9.0 g dl−1, platelet count ≥100 × 109/l, and absolute neutrophil count ≥1.5 × 109/l), hepatic (alanine aminotransferase and aspartate aminotransferase ≤3 × the upper limit of normal [ULN; ≤5 × ULN in patients with liver metastases], and total bilirubin ≤1.5 × ULN), clotting (international normalized ratio ≤ 1.5), and renal function (serum creatinine ≤1.5 × ULN).
Excluded were patients with hepatocellular carcinoma, chronic liver cirrhosis, a history of active coronary artery disease within the previous 6 months (i.e. myocardial infarction, unstable angina, coronary artery bypass graft, or stenting), and/or evidence of uncontrolled symptomatic bradycardia (grade ≥ 2 according to National Cancer Institute [NCI] Common Terminology Criteria for Adverse Events [CTCAE], version 4) or other cardiac arrhythmia or uncontrolled hypertension. Patients who were receiving any known inducer or inhibitor of CYP1A2, CYP2C9, CYP2C19, CYP3A4, or P‐gp, and/or those who had received a nondrug agent or systemic gastric pH modifiers (i.e. ranitidine, proton pump inhibitor, etc.) within the previous 14 days were ineligible.
Treatment protocol
A modified version of the Cooperstown 5 + 1 cocktail was used in this study; the probe for CYP2D6 (dextromethorphan) was omitted because in vitro assessment demonstrated that tivantinib does not inhibit CYP2D6. All patients received study drugs in a fixed sequence, with probe drugs administered twice: once with tivantinib and once without tivantinib. On day 1, patients received simultaneous single oral doses of warfarin 10 mg, vitamin K 5 mg, and caffeine 200 mg with a meal followed in close succession by a single intravenous infusion of midazolam 1.5 mg. Approximately 2 h later, patients received a single oral dose of omeprazole 40 mg. Vitamin K was also given on days 2 and 3 to prevent the anticoagulant effects of warfarin. On day 4, patients received a single oral dose of digoxin 0.25 mg with a meal. Digoxin was administered approximately 72 h after the administration of other CYP probes to avoid the antisecretory effect of omeprazole, which may affect the bioavailability of digoxin. On day 6, all patients started treatment with oral tivantinib 360 mg BID with food, and continued tivantinib through day 15 (total duration of 10 days). The tivantinib dose of 360 mg BID was selected because it is the highest dose being evaluated in clinical studies. After 5 days of treatment with tivantinib, the probe drugs and vitamin K were administered again in the same sequence (warfarin, vitamin K, caffeine, midazolam, and omeprazole on day 11; vitamin K on days 12 and 13; and digoxin on day 14). After completion of the DDI study, patients could continue tivantinib treatment for up to 12 cycles of 28 ± 2 days each. The efficacy results of this extension phase are not reported here.
Blood sampling for PK analyses
Blood sampling times were optimized for each probe drug using the information from the published literature to reduce the number of site visits and volume of blood for collection, and to avoid blood sampling late at night (Table 1). Blood samples were collected by venipuncture using a dipotassium EDTA tube. Plasma was prepared within 45 min of extraction by centrifugation at 1500g for 15 min and stored at −10°C, while whole blood was stored at −20°C until analysis. Blood samples were collected for PK analyses within 60 min before the first doses of warfarin and caffeine and then 5 min (end of midazolam infusion) and 1, 2, 3, 4, 6, and 8 h postdose on days 1 and 11; 60 min before digoxin dosing and 1, 2, 4, 6, and 8 h postdose of digoxin on days 4 and 14; and 60 min before the morning dose of tivantinib and 1, 2, 4, 6, 8, 10, and 12 h after dosing on days 6 and 10. Additional samples were collected for determination of trough tivantinib concentrations before tivantinib dosing on days 14 and 16; predose samples collected on days 4 and 14 were also used to determine S‐warfarin concentrations at 72 h postdose, and predose samples collected on days 6 and 16 were used to determine digoxin concentrations at 48 h postdose and S‐warfarin concentrations at 120 h postdose.
Table 1.
Probe drugs and blood sampling times
| Enzyme of interest | Probe drug | Dose (mg) | Route of administration | Blood sampling time |
|---|---|---|---|---|
| CYP1A2 | Caffeine | 200.00 | PO | Predose and 1, 2, 3, 4, 6, and 8 h postdose |
| CYP2C9 | Warfarin | 10.00 | PO | Predose and 1, 2, 3, 4, 6, 8, 72, and 120 h postdose |
| CYP2C19 | Omeprazole | 40.00 | PO | Predose and 1, 2, 4, and 6 h postdose |
| CYP3A4 | Midazolam | 1.5 | IV | Predose and 0.08, 1, 2, 3, 4, 6, and 8 h postdose |
| P‐gp | Digoxin | 0.25 | PO | Predose and 1, 2, 3, 4, 6, 8, and 48 h postdose |
CYP, cytochrome P450; IV, intravenous; P‐gp, P‐glycoprotein; PO, by mouth
Plasma concentration assays
Plasma samples were assayed for caffeine, S‐warfarin, omeprazole, 5‐hydroxyomeprazole, midazolam, digoxin and tivantinib and its metabolites (HPM4, HPM5, HPM6 and HPM8) by prevalidated, high‐performance liquid chromatography–tandem mass spectrometry (Table S1). Calibration standard data, quality control sample data, incurred sample reproducibility data and chromatograms indicated that assay methods were performed acceptably during the sample analysis. Assay sensitivity and specificity and analyte stability were assessed and successfully validated. All assays were performed within a relative standard deviation of precision and sensitivity of ≤20% and mean accuracy of 80–120% at the lower limit of quantitation.
Blood sampling for pharmacogenomic analyses and genotyping
A blood sample (10 ml) was collected from all patients on day 1 and used for DNA preparation. CYP2C9 and CYP2C19 genotyping was conducted by Cancer Genetics, Inc. (Raleigh, NC, USA), and the genotype data were used to predict metabolizer phenotypes.
PK analyses
The PK parameters included maximum plasma concentration (C max), time to C max (T max), and AUC from time zero to the last quantifiable concentration (AUC last) and from time zero extrapolated to infinity (AUC 0–inf), and were calculated for each probe substrate using SAS 9.1 (SAS Institute, Cary, NC, USA). The ratio of omeprazole to 5‐hydroxyomeprazole was calculated for C max and AUC. PK parameters included AUC 0–12 for tivantinib, and its metabolites (HPM4, HPM5, HPM6 and HPM8) were also calculated after administration of single and multiple doses of tivantinib.
The linear trapezoidal rule was used to calculate AUC. AUC last was calculated for omeprazole and 5‐hydroxyomeprazole when at least three time points with measurable plasma concentrations were available. AUC 0–inf was calculated only if the concentration–time profile exhibited an elimination phase and met the requirements for estimation of the elimination rate constant (R 2 > 0.8, the portion of AUC extrapolated from time to the last quantifiable concentration to infinity did not exceed 20%).
Safety monitoring
Safety was monitored by assessing adverse events, physical examinations, vital signs, electrocardiograms, and clinical laboratory tests. Adverse events were graded according to NCI CTCAE, version 4.
Statistical analyses
Assuming a test‐to‐reference ratio of 1.05, an intra‐patient coefficient of variation of 0.20, a one‐sided type I error rate of 0.05, and a bioequivalence range of 0.80–1.25, a sample size of 20 evaluable patients would provide ≥80% probability to establish equivalence (no drug interaction effect) in C max and AUC between the test treatment (tivantinib in combination with omeprazole, warfarin, caffeine, midazolam or digoxin) and the reference treatment (omeprazole, warfarin, caffeine, midazolam, or digoxin without tivantinib).
The PK‐evaluable set for each drug included all patients who were 100% compliant and took no prohibited concomitant medications. Patients with predose drug concentrations >5% of C max or who were not 100% compliant for a certain drug were excluded from the PK‐evaluable set for that drug. PMs of CYP2C9 were to be excluded from PK analyses of S‐warfarin, and PMs of CYP2C19 were excluded from PK analyses of omeprazole.
For the primary end point statistical analysis, a linear mixed‐effects model was fitted to the log‐transformed PK parameters with treatment (with or without tivantinib) as fixed effects and patient as a random effect. The point estimate of the treatment difference and the corresponding 90% confidence interval (CI) were calculated and antilogged to obtain the point estimate and 90% CI on the linear scale for the ratio of geometric means of the test treatment (probe drug + tivantinib) as compared with the reference treatment (probe drug alone). The absence of a DDI was concluded when the 90% CI of the test‐to‐reference ratio for a PK parameter was within the no effect boundaries of 0.80–1.25. The interpretation of any apparent difference was considered for clinical significance when the 90% CI of the test‐to‐reference ratio for a PK parameter fell outside of the no effect boundaries.
Nomenclature of targets and ligands
Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY 23, and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 24.
Results
Patients
A total of 28 patients were enrolled, received at least one dose of study medication, and had at least one measurable drug concentration (Table 2). Six patients withdrew prematurely from the primary PK study: three (10.7%) because of adverse events, two (7.1%) for other reasons, and one (3.6%) because of progressive disease. One of these patients contributed tivantinib PK data on day 1 before withdrawal, resulting in availability of PK data for 23 patients at this time point; however, only 22 patients completed the primary PK study. There were six PMs of CYP2C9 and no PMs of CYP2C19 in this study.
Table 2.
Demographic characteristics of patients included in the safety analysis set of the primary objective phase
| Characteristic | Safety analysis set (n = 28) |
|---|---|
| Age, years, mean (range) | 61.3 (28.9–84.7) |
| Sex, n (%) | |
| Male | 17 (60.7) |
| Female | 11 (39.3) |
| Race, n (%) | |
| White | 24 (85.7) |
| Black or African American | 2 (7.1) |
| Asian | 1 (3.6) |
| Other | 1 (3.6) |
| Ethnicity, n (%) | |
| Hispanic or Latino | 4 (14.3) |
| Not Hispanic or Latino | 22 (78.6) |
| Unknown | 2 (7.1) |
| ECOG performance status, n (%) | |
| 0 | 6 (21.4) |
| 1 | 20 (71.4) |
| 2 | 2 (7.1) |
| Primary cancer type, n (%) | |
| Colorectal | 5 (17.9) |
| Pancreatic | 3 (10.7) |
| Breast | 2 (7.1) |
| Head and neck | 2 (7.1) |
| Ovarian | 2 (7.1) |
| Lung | 2 (7.1) |
| Prostate | 2 (7.1) |
| Renal | 2 (7.1) |
| Other | 8 (28.6) |
ECOG, Eastern Cooperative Oncology Group
PK effect of tivantinib on probe drugs
Coadministration of tivantinib did not significantly affect the PK of caffeine as evidenced by the geometric least squares (LS) mean ratios (test‐to‐reference) for AUC last (0.97) and C max (1.04) and their 90% CIs, which remained within the predefined equivalence range of 0.80–1.25 (Table 3).
Table 3.
Summary of analysis of variance of probe drug PK parameters to assess the effect of coadministered tivantinib
| PK parameter | N | Geometric LS mean | Test‐to‐reference ratio | 90% CIc (intrapatient CV%) | |
|---|---|---|---|---|---|
| Reference | Test | ||||
| Probe drug alonea | Probe drug + tivantinibb | ||||
| Caffeine (probe for CYP1A2) | |||||
| C max , ng ml −1 | 15 | 4369.4 | 4565.6 | 1.04 | 0.96–1.14 (13.5) |
| AUC last , ng∙hr ml −1 | 15 | 21 220.7 | 20 575.9 | 0.97 | 0.89–1.05 (13.1) |
| S‐Warfarin (probe for CYP2C9) | |||||
| C max , ng ml −1 | 13 | 512.8 | 480.4 | 0.94 | 0.82–1.07 (19.6) |
| AUC last , ng∙hr ml −1 | 13 | 19 418.6 | 17 149.0 | 0.88 | 0.76–1.02 (21.1) |
| Omeprazole (probe for CYP2C19) | |||||
| C max , ng ml −1 | 13 | 866.9 | 774.5 | 0.89 | 0.58–1.38 (68.4) |
| AUC last , ng∙hr ml −1 | 13 | 2173.9 | 1933.7 | 0.89 | 0.60–1.31 (59.7) |
| 5‐hydroxyomeprazole (probe for CYP2C19) | |||||
| C max , ng ml −1 | 13 | 185.0 | 174.0 | 0.94 | 0.72–1.22 (39.0) |
| AUC last , ng∙hr ml −1 | 13 | 503.6 | 513.9 | 1.02 | 0.72–1.45 (54.1) |
| Omeprazole metabolic ratio d | |||||
| AUC last omeprazole to 5‐hydroxyomeprazole | 13 | 4.32 | 3.76 | 0.87 | 0.71–1.08 (30.8) |
| Midazolam (probe for CYP3A4) | |||||
| C max , ng ml −1 | 22 | 74.9 | 66.8 | 0.89 | 0.68–1.17 (55.6) |
| AUC last , ng∙hr ml −1 | 22 | 107.1 | 88.5 | 0.83 | 0.67–1.02 (41.7) |
| Digoxin (probe for P‐glycoprotein) | |||||
| C max , ng ml −1 | 20 | 1.2 | 0.9 | 0.75 | 0.60–0.95 (44.8) |
| AUC last , ng∙hr ml −1 | 20 | 11.5 | 7.9 | 0.69 | 0.51–0.94 (62.0) |
Probe drugs were administered in the absence of tivantinib (day 1 or day 4)
Probe drugs were administered in the presence of tivantinib (steady state, day 11, or day 14)
Lower limit, upper limit of 90% CI
Ratio of AUClast omeprazole to 5‐hydroxyomeprazole
AUC last, area under the concentration–time curve from time 0 to the last quantifiable concentration; CI, confidence interval; C max, maximum plasma concentration; CV, coefficient of variation; CYP, cytochrome P450; LS, least squares; PK, pharmacokinetic
Coadministration of tivantinib reduced exposure to S‐warfarin, midazolam and omeprazole by approximately 12%, 17% and 11%, respectively. In each case, the test‐to‐reference ratio of geometric LS mean AUC last for each analyte was below 1.00. Although the corresponding 90% CIs were not contained within the predefined equivalence range of 0.80–1.25, the upper bound of the corresponding 90% CIs was greater than 1, implying an inconclusive drug interaction (Table 3). In the case of S‐warfarin, the test‐to‐reference ratio of AUC last was 0.88, and the 90% CI was narrow (0.76–1.02), but the lower bound fell just outside of the prespecified range. Similarly, the test‐to‐reference ratio of AUC last for midazolam was 0.83, and the lower bound of the 90% CI (0.67–1.02) fell below the predefined threshold for equivalence. The test‐to‐reference ratio of AUC last for omeprazole (0.89) was similar to that for S‐warfarin, but the 90% CI was broad (0.60–1.31), and both the upper and lower bounds fell outside the prespecified range.
Among the probe drugs evaluated, exposure to digoxin decreased by the greatest amount (31%) after coadministration of tivantinib, and the upper bound of the 90% CI (0.51–0.94) was below 1, suggesting a potential interaction of tivantinib and P‐gp substrates.
Results were mostly similar to those for AUC last when the test‐to‐reference ratio of geometric LS mean C max was calculated for caffeine, omeprazole, midazolam and digoxin (Table 3). In the case of S‐warfarin, the 90% CI remained within the prespecified boundary for equivalence (0.82–1.07).
An overview of the equivalence analysis is provided in Figure 1. The concentration–time profile for each probe drug administered alone or in combination with tivantinib is shown in Figure 2.
Figure 1.
Forest plot of the fold‐change and 90% confidence intervals expressing the effect of tivantinib on the pharmacokinetics of probe drugs for selected cytochrome P450 (CYP) isoenzymes and P‐glycoprotein (P‐gp). Dotted lines represent bioequivalent limits. AUC last, area under the concentration–time curve from time zero to the last quantifiable concentration; C max, maximum plasma concentration
Figure 2.
Plasma concentration–time profile of (A) caffeine, (B) S‐warfarin, (C) omeprazole, (D) 5‐hydroxy omeprazole, (E) midazolam and (F) digoxin after administration of single doses of probe drugs alone or in combination with tivantinib. Arithmetic mean data are shown
Plasma PK profile of tivantinib
PK data for tivantinib and its metabolites were available for 23 patients (Table 4). The plasma concentration–time profile of tivantinib after administration of single and multiple oral doses is shown in Figure 3. Exposure to tivantinib and its metabolites increased with multiple dosing. Arithmetic mean (standard deviation [SD]) C max was 2886.0 (1936.5) ng ml−1 following a single dose and 3201.6 (2603.6) ng ml−1 following multiple doses. Arithmetic mean (SD) AUC 0–12 was 21 000.4 (18 224.7) ng∙hr ml−1 following a single dose and 25 819.3 (28 016.0) ng∙hr ml−1 following multiple doses. The mean accumulation ratio for tivantinib, calculated as the ratio of the arithmetic mean AUC 0–12 after single and multiple doses (day 6/day 10), was 1.15 (1.10 when calculated as the ratio of C max).
Table 4.
Pharmacokinetic parameters for tivantinib
| Parameter | Tivantinib |
|---|---|
| Single dose (day 6) | |
| C max , ng ml −1 a | 2398.4 (69.4) |
| AUC 0–12 , ng∙hr ml −1 a | 15 341.8 (96.4) |
| t max , h b | 4.0 (1.0–8.0) |
| Multiple dose (day 10) | |
| C max , ng ml −1 a | 2453.2 (84.1) |
| AUC 0–12 , ng∙hr ml −1 a | 16 202.3 (124.5) |
| t max , h b | 2.0 (1.0–6.0) |
n = 23.
Geometric mean (% coefficient of variation)
Median (range)
AUC 0–12, area under the concentration–time curve from time 0 to 12 h; C max, maximum plasma concentration; t max, time to C max
Figure 3.
Plasma concentration–time profile of tivantinib after oral administration of a 360 mg twice‐daily dosing. Arithmetic mean data with standard deviation are shown. BID, twice daily
Safety
During the PK evaluation, 22 patients (78.6%) experienced treatment‐emergent adverse events (TEAEs), 14 (50.0%) of whom experienced a tivantinib‐related TEAE and four (14.3%) of whom experienced a tivantinib‐related grade ≥ 3 TEAE. The most frequent tivantinib‐related TEAEs were fatigue in five patients (17.9%) and anaemia, nausea and vomiting in two patients (7.1%) each.
Three patients (10.7%) experienced serious adverse events (SAEs) during the PK study that were related to tivantinib treatment. One patient had nausea and vomiting (related to tivantinib) as well as neutropaenia and a decreased white blood cell count (unrelated to tivantinib), and one patient had anaemia and decreased neutrophil and white blood cell counts (all related to tivantinib). These SAEs caused interruption in tivantinib dosing for both of these patients, resulting in discontinuation from the study. A third patient experienced anaemia related to tivantinib and discontinued treatment because of fatigue. No patients died due to an adverse event during the PK study.
Discussion
This study was designed to evaluate the drug interaction potential of tivantinib with a substrate of CYP1A, CYP2C9, CYP2C19, CYP3A4 or P‐gp at a clinically relevant multiple‐dose regimen in patients with advanced malignancies. If tivantinib inhibited processes mediated by these entities at concentrations similar to those achieved in plasma with doses being evaluated in Phase 3 studies, then one would expect that plasma concentrations of substrates such as the probe drugs used in this study would increase. A mixture of probe drugs for CYP1A2 (caffeine), CYP2C9 (warfarin), CYP2C19 (omeprazole), CYP3A4 (midazolam) and P‐gp (digoxin) were administered alone and subsequently in combination with tivantinib 360 mg BID, and a recommended definition for detection of the potential for DDI was used to objectively assess the data 18.
Collectively, the results of this study show that tivantinib does not alter the systemic exposure of drugs metabolized by CYP1A2, that there is minimal potential to interact with substrates of CYP2C19 and CYP3A4, and that there may be a small interaction between tivantinib and CYP2C9 substrates that is unlikely to be clinically important.
As previously discussed, tivantinib induces CYP3A4 enzyme activity, which is consistent with the slight reduction in plasma levels of the CYP3A4 substrate midazolam when administered with tivantinib in the present study. Similarly, coadministration of tivantinib was associated with a reduction in digoxin AUC last and C max of approximately 31% and 25%, respectively, suggesting that tivantinib is a weak inducer of P‐gp. The CYP3A4 gene and the multidrug resistance 1 (MDR1) gene (which produces P‐gp) are both regulated by the pregnane X receptor, and a large number of drugs and dietary supplements concomitantly induce the expression of CYP3A4 and MDR1 genes 25. Since CYP3A4 is mainly expressed in the liver as well as in the small intestine, this may suggest that the induction of CYP3A4 by tivantinib may potentially decrease the availability of oral drugs that are CYP3A4 substrates 26.
A large number of TKIs have been approved and are in clinical use as anticancer agents in many countries. Since these medications are generally taken on a chronic basis to provide ongoing tumour suppression, there is broad scope for these drugs to be involved in PK DDIs. Almost all TKIs are metabolized via CYP3A4; thus, as a class, interactions involving TKIs and drugs metabolized or inhibited by this pathway have been a focus of attention 19, 27. DDIs are more likely to be clinically significant when drug elimination occurs primarily through a single metabolic pathway, when one of the interacting drugs has an active metabolite and/or a narrow therapeutic window 19. Tivantinib may be somewhat atypical in that it is not metabolized to a large extent by CYP3A4, and the results of this study suggest that it has little potential to alter the metabolism of drugs metabolized via this isoenzyme. Rather, tivantinib is metabolized primarily by CYP2C19, and exposure to the drug varies with host genotype 14, 17. In addition, results from this study suggest that there is little potential for tivantinib to alter the PK of substrates of other CYP isozymes (i.e. CYP1A2, CYP2C9 and CYP2C19). Many TKIs and cytotoxic drugs, such as vinblastine and docetaxel, are substrates for efflux transporters such as P‐gp 27; thus, our finding that exposure to digoxin (a P‐gp substrate) is slightly decreased is reassuring and suggests that tivantinib has little potential to reduce exposure to drugs that are substrates of P‐gp.
The PK profile of oral tivantinib in the present study was consistent with previous reports in patients with cancer receiving 360 mg BID in a crystalline formulation. Considerable inter‐individual variability and less than proportional increases in plasma levels have been reported in the PK of tivantinib 12.
Tivantinib was well tolerated in the present study, with the spectrum and frequency of TEAEs being consistent with those reported previously. Of note, bradycardia has been infrequently reported in patients treated with tivantinib 8, 9, 28 but was not reported in any patients in the present study, including during treatment with digoxin, which is associated with bradycardia.
Limitations of this study include the small number of patients and the lack of probes for CYP2B6 and CYP2C8 that have not been validated as probe substrates in the DDI cocktail. The number of evaluable subjects for some probes further decreased at end because of exclusion of patients that were not evaluable for PK analysis, which may result in low statistical power. As such, the clinical significance of the DDI effect was discussed with caution. The potential of clinically significant DDIs between tivantinib and the substrates for CYP2B6 and CYP2C8 is considered to be unlikely because no to minimal effect of tivantinib was observed on the PK of substrates for other CYPs with similar Ki values.
Strengths of the present study include the use of a cocktail mixture of CYP/P‐gp probe drugs, at clinically relevant doses, to assess the drug interaction potential of tivantinib in patients with cancer. This approach has been previously used for nononcology drugs in healthy subjects but rarely used for oncology drugs in patients with cancer 29, 30. Use of the cocktail may decrease intra‐patient variability in CYP enzyme activity. The involvement of patients rather than healthy subjects provides data in the target population at the clinically relevant exposure of the interacting drug. The activity of certain CYP isoenzymes may be altered in patients with cancer 31, 32. These patients are also more likely to have kidney hepatic dysfunction, low albumin levels, and comorbid medical conditions, and to be taking multiple medications – all of which may predispose them to adverse events or DDIs 27. This study allowed an opportunity to evaluate DDI potential in a clinical setting. In addition, this study instituted an optimized blood‐sampling time schedule for each probe drug that reduced the number of site visits as well as the volume of blood for collection, and helped avoid late‐night blood sampling. Evaluations were conducted under steady‐state conditions after administration of tivantinib 360 mg BID. This design allowed for the simultaneous evaluation of four major CYP enzymes and P‐gp, which are implicated in many clinically significant DDIs.
Conclusions
Overall, the data suggest that tivantinib 360 mg BID has either a minimal or no effect on the PK of CYP1A2, CYP2C9, CYP2C19 and CYP3A4. It is likely that the DDI effect is of minimal clinical significance when CYP1A2, CYP2C9, CYP2C19 and CYP3A4 substrates are concomitantly administered with tivantinib in patients with cancer. Close monitoring might be necessary in terms of loss of efficacy for P‐gp substrates with a narrow therapeutic window, as coadministration of tivantinib would decrease systemic exposure to the drug.
Competing Interests
M.T., R.G., Y.W. and H.Z. are employees of Daiichi Sankyo, Inc. K.P.P. has received research funding to South Texas Accelerated Research Therapeutics (START) for the conduct of this clinical trial. J.H.S. has received research funding to Duke University Medical Center. I.P. has nothing to disclose.
The authors would like to acknowledge the bioanalysis oversight by Ling He and Miho Kazui. The study and development of the manuscript were supported by Daiichi Sankyo, Inc. Jessica Deckman and Robert Schupp of inScience Communications, Springer Healthcare, and Blair Jarvis on behalf of inScience Communications, Springer Healthcare, provided medical writing support funded by Daiichi Sankyo, Inc.
Contributors
M.T. analysed and interpreted data, wrote the first draft, and reviewed and revised the manuscript. K.P.P. designed and performed research and reviewed and revised the manuscript J.H.S. performed research and reviewed and revised the manuscript. I.P. performed research and critically edited, reviewed, and revised the manuscript. R.G. acquired data and performed clinical operations. Y.W. analysed and interpreted data. H.Z. conceptualized and designed the study, acquired data, analysed and interpreted data, and reviewed and revised the manuscript.
Supporting information
Table S1 Characteristics of bioanalytical assays
Tachibana, M. , Papadopoulos, K. P. , Strickler, J. H. , Puzanov, I. , Gajee, R. , Wang, Y. , and Zahir, H. (2018) Evaluation of the pharmacokinetic drug interaction potential of tivantinib (ARQ 197) using cocktail probes in patients with advanced solid tumours. Br J Clin Pharmacol, 84: 112–121. doi: 10.1111/bcp.13424.
PI Statement: Kyriakos P. Papadopoulos, John H. Strickler, and Igor Puzanov were the principal investigators for this study.
References
- 1. Gherardi E, Birchmeier W, Birchmeier C, Vande Woude G. Targeting MET in cancer: rationale and progress. Nat Rev Cancer 2012; 12: 89–103. [DOI] [PubMed] [Google Scholar]
- 2. Munshi N, Jeay S, Li Y, Chen CR, France DS, Ashwell MA, et al ARQ 197, a novel and selective inhibitor of the human c‐Met receptor tyrosine kinase with antitumor activity. Mol Cancer Ther 2010; 9: 1544–1553. [DOI] [PubMed] [Google Scholar]
- 3. Calles A, Kwiatkowski N, Cammarata BK, Ercan D, Gray NS, Janne PA. Tivantinib (ARQ 197) efficacy is independent of MET inhibition in non‐small‐cell lung cancer cell lines. Mol Oncol 2015; 9: 260–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Katayama R, Aoyama A, Yamori T, Qi J, Oh‐hara T, Song Y, et al Cytotoxic activity of tivantinib (ARQ 197) is not due solely to c‐MET inhibition. Cancer Res 2013; 73: 3087–3096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Basilico C, Pennacchietti S, Vigna E, Chiriaco C, Arena S, Bardelli A, et al Tivantinib (ARQ197) displays cytotoxic activity that is independent of its ability to bind MET. Clin Cancer Res 2013; 19: 2381–2392. [DOI] [PubMed] [Google Scholar]
- 6. Lu S, Torok HP, Gallmeier E, Kolligs FT, Rizzani A, Arena S, et al Tivantinib (ARQ 197) affects the apoptotic and proliferative machinery downstream of c‐MET: role of Mcl‐1, Bcl‐xl and Cyclin B1. Oncotarget 2015; 6: 22167–22178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Adjei AA, Schwartz B, Garmey E. Early clinical development of ARQ 197, a selective, non‐ATP‐competitive inhibitor targeting MET tyrosine kinase for the treatment of advanced cancers. Oncologist 2011; 16: 788–799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Santoro A, Rimassa L, Borbath I, Daniele B, Salvagni S, Van Laethem JL, et al Tivantinib for second‐line treatment of advanced hepatocellular carcinoma: a randomised, placebo‐controlled phase 2 study. Lancet Oncol 2013; 14: 55–63. [DOI] [PubMed] [Google Scholar]
- 9. Scagliotti G, von Pawel J, Novello S, Ramlau R, Favaretto A, Barlesi F, et al Phase III multinational, randomized, double‐blind, placebo‐controlled study of tivantinib (ARQ 197) plus erlotinib versus erlotinib alone in previously treated patients with locally advanced or metastatic nonsquamous non‐small‐cell lung cancer. J Clin Oncol 2015; 33: 2667–2674. [DOI] [PubMed] [Google Scholar]
- 10. Rimassa L, Porta C, Borbath I, Bruno D, Finn RS, Raoul J‐L, et al Tivantinib in MET‐high hepatocellular carcinoma patients and the ongoing phase III clinical trial. Hepatic Oncol 2014; 1: 181–188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Murai T, Takakusa H, Nakai D, Kamiyama E, Taira T, Kimura T, et al Metabolism and disposition of [(14)C]tivantinib after oral administration to humans, dogs and rats. Xenobiotica 2014; 44: 996–1008. [DOI] [PubMed] [Google Scholar]
- 12. Rosen LS, Senzer N, Mekhail T, Ganapathi R, Chai F, Savage RE, et al A phase I dose‐escalation study of tivantinib (ARQ 197) in adult patients with metastatic solid tumors. Clin Cancer Res 2011; 17: 7754–7764. [DOI] [PubMed] [Google Scholar]
- 13. Yap TA, Olmos D, Brunetto AT, Tunariu N, Barriuso J, Riisnaes R, et al Phase I trial of a selective c‐MET inhibitor ARQ 197 incorporating proof of mechanism pharmacodynamic studies. J Clin Oncol 2011; 29: 1271–1279. [DOI] [PubMed] [Google Scholar]
- 14. Yamamoto N, Murakami H, Nishina T, Hirashima T, Sugio K, Muro K, et al The effect of CYP2C19 polymorphism on the safety, tolerability, and pharmacokinetics of tivantinib (ARQ 197): results from a phase I trial in advanced solid tumors. Ann Oncol 2013; 24: 1653–1659. [DOI] [PubMed] [Google Scholar]
- 15. Okusaka T, Aramaki T, Inaba Y, Nakamura S, Morimoto M, Moriguchi M, et al Phase I study of tivantinib in Japanese patients with advanced hepatocellular carcinoma: distinctive pharmacokinetic profiles from other solid tumors. Cancer Sci 2015; 106: 611–617. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Geller JI, Perentesis JP, Liu X, Minard CG, Kudgus RA, Reid JM, et al A phase 1 study of the c‐Met inhibitor, tivantinib (ARQ197) in children with relapsed or refractory solid tumors: a Children's Oncology Group study phase 1 and pilot consortium trial (ADVL1111). Pediatr Blood Cancer 2017; 64: e26565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Yamamoto N, Murakami H, Hayashi H, Fujisaka Y, Hirashima T, Takeda K, et al CYP2C19 genotype‐based phase I studies of a c‐Met inhibitor tivantinib in combination with erlotinib, in advanced/metastatic non‐small cell lung cancer. Br J Cancer 2013; 109: 2803–2809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Food and Drug Administration . Guidance for Industry. Drug interaction studies – study design, data analysis, implications for dosing, and labeling recommendations. Rockville, MD: US Department of Health and Human Services CfDEaRC, 2012; 79. [Google Scholar]
- 19. Teo YL, Ho HK, Chan A. Metabolism‐related pharmacokinetic drug‐drug interactions with tyrosine kinase inhibitors: current understanding, challenges and recommendations. Br J Clin Pharmacol 2015; 79: 241–253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. van Leeuwen RW, van Gelder T, Mathijssen RH, Jansman FG. Drug‐drug interactions with tyrosine‐kinase inhibitors: a clinical perspective. Lancet Oncol 2014; 15: e315–e326. [DOI] [PubMed] [Google Scholar]
- 21. Turpault S, Brian W, Van Horn R, Santoni A, Poitiers F, Donazzolo Y, et al Pharmacokinetic assessment of a five‐probe cocktail for CYPs 1A2, 2C9, 2C19, 2D6 and 3A. Br J Clin Pharmacol 2009; 68: 928–935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Chainuvati S, Nafziger AN, Leeder JS, Gaedigk A, Kearns GL, Sellers E, et al Combined phenotypic assessment of cytochrome p450 1A2, 2C9, 2C19, 2D6, and 3A, N‐acetyltransferase‐2, and xanthine oxidase activities with the “Cooperstown 5+1 cocktail”. Clin Pharmacol Ther 2003; 74: 437–447. [DOI] [PubMed] [Google Scholar]
- 23. Southan C, Sharman JL, Benson HE, Faccenda E, Pawson AJ, Alexander SPH, et al The IUPHAR/BPS Guide to PHARMACOLOGY in 2016: towards curated quantitative interactions between 1300 protein targets and 6000 ligands. Nucl Acids Res 2016; 44: D1054–D1068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Alexander SPH, Fabbro D, Kelly E, Marrion N, Peters JA, Benson HE, et al The Concise Guide to PHARMACOLOGY 2015/16: Enzymes. Br J Pharmacol 2015; 172: 6024–6109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Rathod V, Jain S, Nandekar P, Sangamwar AT. Human pregnane X receptor: a novel target for anticancer drug development. Drug Discov Today 2014; 19: 63–70. [DOI] [PubMed] [Google Scholar]
- 26. Watkins PB. The barrier function of CYP3A4 and P‐glycoprotein in the small bowel. Adv Drug Deliv Rev 1997; 27: 161–170. [DOI] [PubMed] [Google Scholar]
- 27. Shao J, Markowitz JS, Bei D, An G. Enzyme‐ and transporter‐mediated drug interactions with small molecule tyrosine kinase inhibitors. J Pharm Sci 2014; 103: 3810–3833. [DOI] [PubMed] [Google Scholar]
- 28. Monk P, Liu G, Stadler WM, Geyer SM, Sexton JL, Wright JJ, et al Phase II randomized, double‐blind, placebo‐controlled study of tivantinib in men with asymptomatic or minimally symptomatic metastatic castrate‐resistant prostate cancer (mCRPC). J Clin Oncol 2015; 33 (7 suppl): 146. [Google Scholar]
- 29. Goh BC, Reddy NJ, Dandamudi UB, Laubscher KH, Peckham T, Hodge JP, et al An evaluation of the drug interaction potential of pazopanib, an oral vascular endothelial growth factor receptor tyrosine kinase inhibitor, using a modified Cooperstown 5+1 cocktail in patients with advanced solid tumors. Clin Pharmacol Ther 2010; 88: 652–659. [DOI] [PubMed] [Google Scholar]
- 30. Flaherty KT, Lathia C, Frye RF, Schuchter L, Redlinger M, Rosen M, et al Interaction of sorafenib and cytochrome P450 isoenzymes in patients with advanced melanoma: a phase I/II pharmacokinetic interaction study. Cancer Chemother Pharmacol 2011; 68: 1111–1118. [DOI] [PubMed] [Google Scholar]
- 31. Rivory LP, Slaviero KA, Clarke SJ. Hepatic cytochrome P450 3A drug metabolism is reduced in cancer patients who have an acute‐phase response. Br J Cancer 2002; 87: 277–280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Helsby NA, Lo WY, Sharples K, Riley G, Murray M, Spells K, et al CYP2C19 pharmacogenetics in advanced cancer: compromised function independent of genotype. Br J Cancer 2008; 99: 1251–1255. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1 Characteristics of bioanalytical assays