Evaluation of the Cytochrome P450 3A and P‐glycoprotein Drug‐Drug Interaction Potential of Futibatinib

Futibatinib, a selective, irreversible fibroblast growth factor receptor 1–4 inhibitor, is being investigated for tumors harboring FGFR aberrations and was recently approved for the treatment of FGFR2 fusion/rearrangement‐positive intrahepatic cholangiocarcinoma. In vitro studies identified cytochrome P450 (CYP) 3A as the major CYP isoform in futibatinib metabolism and indicated that futibatinib is likely a P‐glycoprotein (P‐gp) substrate and inhibitor. Futibatinib also showed time‐dependent inhibition of CYP3A in vitro. Phase I studies investigated the drug‐drug interactions of futibatinib with itraconazole (a dual P‐gp and strong CYP3A inhibitor), rifampin (a dual P‐gp and strong CYP3A inducer), or midazolam (a sensitive CYP3A substrate) in healthy adult participants. Compared with futibatinib alone, coadministration of futibatinib with itraconazole increased futibatinib mean peak plasma concentration and area under the plasma concentration–time curve by 51% and 41%, respectively, and coadministration of futibatinib with rifampin lowered futibatinib mean peak plasma concentration and area under the plasma concentration–time curve by 53% and 64%, respectively. Coadministration of midazolam with futibatinib had no effect on midazolam pharmacokinetics compared with midazolam administered alone. These findings suggest that concomitant use of dual P‐gp and strong CYP3A inhibitors/inducers with futibatinib should be avoided, but futibatinib can be concomitantly administered with other drugs metabolized by CYP3A. Drug‐drug interaction studies with P‐gp–specific substrates and inhibitors are planned.

3][4][5] In a phase I dose-escalation study (NCT02052778), maximum plasma concentrations (C max ) of futibatinib and area under the plasma concentration-time curve (AUC) were found to increase in a dose-proportional manner up to 80 mg of futibatinib orally administered in patients with advanced solid tumors refractory to standard therapies. 2ased on safety, PK, and pharmacodynamic data, 20 mg once daily was determined to be the recommended phase II dose.
Futibatinib showed antitumor activity and tolerability in patients with various types of FGFR-aberrant solid tumors in phase I studies. 2,5,6In a phase II study, futibatinib 20 mg once daily demonstrated durable responses (objective response rate, 41.7%; median duration of response, 9.7 months) and promising survival (median progression-free survival, 9.0 months; median overall survival, 21.7 months) in patients with advanced, previously treated intrahepatic cholangiocarcinoma (iCCA) harboring FGFR2 fusions or rearrangements. 7,8Futibatinib is approved and used clinically for patients with previously treated, unresectable, locally advanced or metastatic iCCA harboring FGFR2 fusions or rearrangements.6][7][8] The most common treatment-emergent adverse event (AE) was hyperphosphatemia, an on-target off-tumor effect due to inhibition of FGFR1. 2,6,7,9everal phase I studies were conducted in healthy participants to further evaluate the impact of extrinsic factors on futibatinib PK.One such study showed that the consumption of a high-fat, high-calorie meal had no clinically meaningful impact on the overall futibatinib exposure. 3Another phase I study showed that coadministration of the proton pump inhibitor lansoprazole did not affect futibatinib PK. 3 An openlabel, phase I mass balance study in healthy participants showed that futibatinib was the major circulating component in plasma. 4Circulating metabolites in plasma were cysteinylglycine-conjugated futibatinib, glucuronide of mono-hydroxyl futibatinib, and cysteine-conjugated futibatinib.After a single dose of [14C]-futibatinib, most of the radioactivity (64% of a total 70%) was recovered in the feces, with the main metabolites identified in feces being derivatives of desmethyl futibatinib, monohydroxyl futibatinib, and glutathione-conjugated futibatinib.The excretion of unchanged futibatinib was negligible, and renal excretion was only a minor excretion pathway (6% of the radioactivity was recovered in the urine). 4These results suggest that both oxidation and glutathione conjugation are involved in futibatinib metabolism.In line with these findings, recent reports have shown extrahepatic metabolism of covalent tyrosine kinase inhibitors similar to futibatinib possessing a Michael acceptor, such as an acrylamide moiety, via nonenzymatic and glutathione S-transferase (GST)-mediated glutathione conjugation. 10ytochrome P450 (CYP) 3A forms a major class of drug-metabolizing enzymes mediating oxidation reactions, and many anticancer drugs are CYP3A substrates.Some but not all drugs metabolized by CYP3A are also transported by the efflux drug transporter Pglycoprotein (P-gp). 11,12Results of a recent in vitro study showed that futibatinib inactivates both predominant CYP3A isoforms (CYP3A4 and CYP3A5) in a time-, concentration-, and cofactor-dependent manner. 13o investigate potential interactions between futibatinib and substrates or modulators of CYP3A and Pgp, we conducted a series of in vitro studies and phase I drug-drug interaction (DDI) studies, whose key findings are presented here.

In Vitro Studies
In Vitro CYP Phenotyping The objective of this study was to identify the metabolizing enzymes of futibatinib. 14,15Details of all relevant reagents are provided in Table S1.First, a pilot experiment was performed by incubating futibatinib with recombinant human cytochrome P450 (rhCYP) enzymes (CYP1A2, 2B6, 2C8, 2C9, 2C19, 2D6, 3A4, and 3A5) or control microsomes fortified with the nicotinamide adenine dinucleotide phosphate reduced form generating system (nicotinamide adenine dinucleotide phosphate reduced from generating system [NADPHgs]) for 60 minutes.Following this pilot, confirmatory experiments were conducted by incubating futibatinib with the specific rhCYP enzymes.Upon confirmation, the contribution of each enzyme to futibatinib metabolism was further evaluated by incubating futibatinib with human liver microsomes in the presence of specific CYP inhibitors.The relative contributions of CYP and GST enzymes were evaluated using the human liver S9 fraction in the presence of a pan-CYP inhibitor (a cocktail of 1-aminobenzotriazole and tienilic acid) or a pan-GST inhibitor (ethacrynic acid).
The incubation mixtures for the reactions with rhCYP enzymes contained 100 mmol/L phosphate buffer (pH 7.4), 5 mmol/L of MgCl 2 , NADPHgs, 0.5 μmol/L (209 ng/mL) futibatinib, 20 pmol/mL rhCYP enzymes, and control microsomes (0.3 mg protein/mL as total rhCYP and control microsomal protein concentration).The reactions with human liver microsomes or S9 fraction were set up similarly but contained 0.5 mg protein/mL human liver microsomes or S9 fraction instead of the rhCYP enzymes.After designated incubation times (60 minutes in the pilot experiment; 0, 15, 30, 45, and 60 minutes for assay with rhCYP enzymes and 0, 30, 60, and 90 minutes for assay with human liver microsomes and S9 fraction in confirmatory experiments), the mixture was filtered, with the final filtrate analyzed using liquid chromatographytandem mass spectrometry (LC-MS/MS).Experiments were performed in duplicate or triplicate.
In Vitro Assessment of P-gp Substrate and P-gp Inhibition In vitro transcellular transport assays were used to evaluate interactions between futibatinib and P-gp transporters. 16P-gp/LLC-PK1 cells (cell lines overexpressing the human P-gp transporter) and mock/LLC-PK1 cells were preincubated in a CO 2 incubator (5% CO 2 ) at 37°C for 10 minutes with transport buffer.Transcellular transports of substrate or futibatinib were initiated with a prewarmed substrate solution (0.25 vol% dimethyl sulfoxide) or a transport buffer containing futibatinib, which were added into each well of the insert and receiver plates to evaluate the transcellular transports in apical (A)-to-basolateral (B) and B-to-A directions.Transport buffer was added to the opposite sides (receiver and insert plates).After incubation for 2 hours, the number of substrates transported across the cells was determined using LC-MS/MS or liquid scintillation counting.
To evaluate the effect of futibatinib on the transcellular transport of a typical substrate, [3H]-verapamil, this experiment was performed similarly, but with substrate solutions containing futibatinib added into the insert and receiver plate wells. 17The following concentrations of futibatinib were used based on the maximum soluble concentration and C max values observed in phase I studies: 2,5 0.03-60 μmol/L (12.6-25,107 ng/mL) for the P-gp inhibition assay; and 0.3-30 μmol/L (126-12,554 ng/mL) for the concentration-dependent transcellular substrate assay.Experiments were performed in triplicate.
In Vitro CYP Inhibition The ability of futibatinib to inhibit CYP enzymes in vitro was evaluated in human liver microsomes. 18,19Human liver microsomes were incubated with standard marker activity substrates and NADPHgs with or without futibatinib to evaluate the IC 50 for CYP enzymes.To evaluate futibatinib as a time-dependent inhibitor of CYP activity, futibatinib was preincubated with human liver microsomes and NADPHgs for 30 minutes before incubating with marker substrates and residual activity was compared with or without preincubation.Known reversible and time-dependent inhibitors of CYP enzymes were included as positive controls (Table S2).At the end of each incubation period, termination solvent was added, and the final incubation mixture was filtered to remove microsomal protein.To calculate time-dependent inhibition kinetic parameters (inhibitor concentration causing half-maximal inactivation and maximal inactivation rate constant) after different preincubation times, aliquots of the enzyme-inhibitor mix were diluted into the substrate solution, and the remaining enzyme activity was determined.Experiments were performed in triplicate.

In Vitro Analysis Methods
Concentrations of futibatinib and metabolites of typical CYP probes in the in vitro assays were determined using LC-MS/MS.In vitro samples were treated with organic solvent, including corresponding internal standards (IS; Table S2), and processed.After chromatographic separation on a C18 reverse-phase column with the mobile phase (0.1% formic acid in acetonitrile for P-gp assays; and 0.1% formic acid in acetonitrile/methanol [1/1, v/v] for CYP assays), the compounds were detected with a triple quadrupole mass spectrometer using electrospray ionization.Positive ions were monitored in the multiple reaction mon-itoring (MRM) mode of the following transitions: m/z 419.2 → 296.2 for futibatinib and m/z 425.2 → 302.2 for the IS (d 6 -futibatinib).The quantitative concentration range of futibatinib was from 10 to 1000 nM (4.18-418 ng/mL) for CYP phenotyping assays and from 1 to 1000 nM (0.418-418 ng/mL) for P-gp assays.MRM transitions and the quantitative concentration ranges of the CYP probe metabolites are provided in Table S3.Quantitation was performed using a weighted linear regression analysis (1/concentration2) of peak area ratios of futibatinib/metabolites and the corresponding IS.The radioactivity of [3H]-verapamil was determined by mixing transport buffer containing [3H]-verapamil with a liquid scintillation cocktail in a plastic vial and then counting radioactivity for 1 minute using a scintillation counter.Total radioactivity was calculated by subtracting background of transport buffer alone.

Phase I Study Designs
Futibatinib and Itraconazole or Rifampin DDI Study Design Participants were healthy nonsmoking adults (including women of nonchildbearing potential), 18-55 years of age, with a body mass index 18.0-32.0kg/m 2 .Key exclusion criteria were history of alcoholism or drug abuse within 2 years before the first dosing, use of drugs known to be significant inducers of CYP3A4 enzymes and/or P-gp (including St. John's wort) within 28 days or during the study, and participation in another clinical study within 30 days before the first dosing.
All study medications were administered orally.After an overnight fast (≥10 hours), participants first received a single dose of 20-mg futibatinib in period 1.In period 2, participants received multiple doses of itraconazole 200 mg once daily (for 6 days) or rifampin 600 mg once daily (for 9 days) along with a single dose of 20 mg futibatinib administered on day 5 or 8, respectively.][22] Dose and timing of rifampin were selected to maximize its induction of CYP3A activity in vivo. 23There was a washout period of at least 2 days between futibatinib dosing in period 1 and the first dose of itraconazole or rifampin in period 2. Blood samples for PK analysis were collected before dosing and through 48 hours after futibatinib dosing.Futibatinib plasma concentrations were determined using LC-MS/MS methods as previously described (Figure 1). 3 The analytical range (lower limit of quantitation to upper limit of quantitation) for futibatinib was 0.50-250 ng/mL.
Futibatinib and Midazolam DDI Study Design All study medications were administered orally.Eligibility criteria for participants were the same as for the previously described study.Participants received a single dose of 2-mg midazolam in period 1 and a dose of  6P2, days 1-6 in period 2; D1-7P2, days 1-7 in period 2; D1-9P2, days 1-9 in period 2; D1P1, day 1 in period 1; D5P2, day 5 in period 2; D7P2, day 7 in period 2; D8P2, day 8 in period 2; PK, pharmacokinetic; QD, once daily.20-mg futibatinib once daily for 7 consecutive days before receiving a single dose of 2-mg midazolam in period 2 (Figure 1).There was a washout period of at least 1 day between midazolam dosing in period 1 and the first dose of futibatinib in period 2. Blood samples for plasma midazolam and 1-OH-midazolam concentrations were collected before dosing through 24 hours after dosing.Blood samples for plasma futibatinib concentrations were collected before dosing and through 12 hours after dosing.Plasma concentrations of midazolam, 1-OH-midazolam, and futibatinib were determined using the LC-MS/MS method.Briefly, midazolam and d4-midazolam (IS), and 1-OH-midazolam and d4-1-OH-midazolam (IS) were extracted by liquid-liquid extraction from plasma treated with βglucuronidase.After chromatographic separation on a diphenyl column with 200 mM ammonium formate in water/methanol as the mobile phase, the compounds were detected with a triple quadrupole mass spectrometer using an electrospray ionization.Positive ions were monitored in the MRM mode of the following transitions: m/z 326.3 → 291.3 for midazolam, m/z 330.3 → 295.3 for the midazolam IS, m/z 342.1 → 203.1 for 1-OH-midazolam, and m/z 346.1 → 328.1 for the 1-OH-midazolam IS; no significant interference was detected.Quantitation was performed using a weighted linear regression analysis (1/concentration 2 ) of peak area ratios of the compounds and the corresponding IS.During the method validation, the within-run accuracy (%bias) and precision (percent coefficient of variation [%CV]) ranged from −3.0% to 8.0% and from 0.8% to 5.4%, respectively, for midazolam, and ranged from −3.3% to 9.0% and from 0.3% to 5.7%, respectively, for 1-OH-midazolam.The between-run accuracy (%bias) and precision (%CV) ranged from −1.3% to 6.0% and from 1.6% to 3.5%, respectively, for midazolam, and ranged from −1.3% to 3.0% and from 1.9% to 5.9%, respectively, for 1-OH-midazolam.The analytical range (lower limit of quantitation to upper limit of quantitation) was 0.10-40.0ng/mL for midazolam, 0.10-20.0ng/mL for 1-OH-midazolam, and 0.50-250 ng/mL for futibatinib.

Statistical Analysis
The futibatinib PK parameters (Table S4) were derived by noncompartmental analysis using Phoenix Win-Nonlin version 7.0 (Certara, Princeton, NJ).A linear mixed-effect analysis of the variance model was used to compare C max , AUC from time 0 extrapolated to infinity (AUC 0-inf ) and AUC from time 0 to the time of the last observed/measured non-0 concentration (AUC 0-t ) between futibatinib administered alone and in combination with itraconazole or rifampin, and between midazolam administered alone and in combination with futibatinib.Time to C max (t max ) was determined using nonparametric analyses.Descriptive statistics were used to summarize AEs, laboratory values, vital signs, and electrocardiogram parameters.
For the DDI study between futibatinib and itraconazole or rifampin, a sample size of 16 participants in each study part was considered sufficient to provide a reliable estimate of the magnitude and variability of any inhibitory or induction interactions by either itraconazole or rifampin, respectively, if interactions exist.To account for potential dropouts, 20 participants were enrolled into each study part, and each participant participated in only 1 study part.For the DDI study between futibatinib and midazolam, a sample size of 22 was determined as being sufficient to confirm the lack of DDIs with a power of at least 80% and an alpha error of 5%. 24wenty-four participants were enrolled in the study to account for possible dropouts.

Compliance With Ethical Standards
Protocols for both studies were approved by an institutional review board (Advarra, Inc., Columbia, MD).The studies were conducted in compliance with Good Clinical Practice Guidelines, principles set forth in the Declaration of Helsinki, the International Conference on Harmonization, and the ethical requirements referred to in the European Union directive 2001/20/EC.All study participants provided written informed consent prior to any study-related procedures.
In Vitro CYP Inhibition The CYP inhibitory potential of futibatinib was evaluated in human liver microsomes from a pool of 50 human individuals.In human liver microsomes, futibatinib reversibly inhibited CYP2C8, CYP2C9, and CYP2C19 (IC 50 , 8.1 μmol/L [3389 ng/mL], 23.9 μmol/L [10,001 ng/mL], and 26.5 μmol/L [11,089 ng/mL], respectively), but not CYP1A2, CYP2B6, CYP2D6, or CYP3A (IC 50 >50 μmol/L [>20,923 ng/mL]).However, the IC 50 values were calculable for CYP3A (based on 1-OH-midazolam and 6ß-hydroxytestosterone) in timedependent inhibition assays, indicating that futibatinib could potentially inhibit CYP3A in a time-dependent manner.There was no evidence of time-dependent inhibition of futibatinib toward CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, and CYP2D6.The concentration causing half-maximal inactivation and maximal inactivation rate constant for inactivation of CYP3A by futibatinib in human liver microsomes was estimated to be 24.6 μmol/L (10,294 ng/mL) and 0.0766 min, respectively, for 1-OH-midazolam, and 62.4 μmol/L (26,111 ng/mL) and 0.136 min for 6ß-hydroxytestosterone (Table 3).Using basic model analyses, 26 the clinical inhibition potential of futibatinib on CYP3A enzymes was assessed.In the presence and absence of futibatinib, the predicted ratio (R 2 ) of the AUC of midazolam was estimated to be 16.45 and the R 2 of testosterone was estimated to be 12.26.As the R 2 value was more than 10-fold higher than the cutoff (R 2 < 1.25), these data suggested that potential in vivo   k deg is the apparent first-order degradation rate constant of the hepatic CYP3A enzymes (0.00032 min −1 ).K I,u is the unbound inhibitor concentration causing half-maximal inactivation and calculated according to the following equation: K I,u = K I × f u,mic , where K I is the inhibitor concentration causing half-maximal inactivation, and f u,mic that is the unbound fraction in microsomes (0.57 at 1 mg/mL microsomal protein) was estimated according to the following equation: where C is the microsomal protein concentration used in the in vitro study for the IC 50 determination for reversible inhibition.I max,u is the unbound C max of futibatinib.K inact is the maximal inactivation rate constant.
DDIs between futibatinib and CYP3A substrates from time-dependent inhibition were considered strong.Similar model analyses ruled out potential in vivo DDIs with CYP2C substrates at clinically relevant concentrations of futibatinib.

Phase I DDI Studies
Participant Demographics and Disposition In the study examining DDIs between futibatinib and itraconazole (a dual P-gp and strong CYP3A inhibitor) or rifampin (a dual P-gp and strong CYP3A inducer), 40 participants were enrolled, of whom 20 received futibatinib and itraconazole (part 1) and 20 received futibatinib and rifampin (part 2).All participants completed the study and were included in the analyses.Most participants were White (85%) and women (55%), and mean age was 40.5 years (Table S5).In the study examining DDIs between futibatinib and midazolam, 24 healthy adults were enrolled, all of whom received treatment and completed the study.Most participants were White (67%) and men (92%), and mean age was 39.2 years (Table S5).
Effect of Itraconazole on Futibatinib PKA plot of mean futibatinib plasma concentration over time with or without itraconazole is shown in Figure 2A  and B. Following the administration of a single dose of futibatinib, futibatinib plasma concentrations increased, reaching a median t max by 1.5 hours (Figure 2A and B; Table 4).Coadministration of futibatinib and itraconazole resulted in higher peak plasma futibatinib concentrations (Figure 2A and B) and earlier futibatinib t max than when futibatinib was administered alone (median difference, −0.7 h; P = .0071;Table 4).Coadministration of multiple oral doses of itraconazole increased futibatinib exposure (Table 4), although the apparent elimination phase of futibatinib was not markedly affected (Figure 2A and B).

Effect of Rifampin on Futibatinib PK A plot of mean futibatinib plasma concentrations over time with or without coadministration of rifampin is shown in
following coadministration with rifampin, shortening the mean apparent first-order terminal elimination half-life (t 1/2 ) of futibatinib from 2.8 to 2.0 hours (Table 4).

Effect of Futibatinib on the PK of Midazolam and 1-OH-Midazolam
A plot of mean midazolam plasma concentrations over time with or without coadministration of futibatinib is shown in Figure 3.The effect of multiple oral doses of futibatinib on the single oral dose PK of midazolam is provided in Table 5. Coadministration of futibatinib and midazolam resulted in midazolam plasma concentrations comparable with those observed with midazolam alone (Table 5).The 90% confidence intervals for total and peak exposure following the coadministration of midazolam and futibatinib were 80%-104% and 84%-106%, respectively, compared with midazolam alone.As the 90% confidence intervals for systemic exposure ratios were within the bioequivalence range of 80%-125%, this result indicated that futibatinib is not an inhibitor of CYP3A enzymes.
Compared with midazolam alone, the coadministration of futibatinib and midazolam resulted in a small t max change (median difference, -0.2 hours; P = .0013;Table 5), which was not clinically meaningful.The t 1/2 and CL/F of midazolam were similar when midazolam was administered alone or with futibatinib (Table 5).Similarly, coadministration of futibatinib 20 mg once daily did not meaningfully impact 1-OH-midazolam PK parameters following a single oral dose of 2-mg midazolam.Exposure to 1-OH-midazolam was also comparable between midazolam and futibatinib coadministered and midazolam alone (Figure 3 and Table S6).Following a single oral dose of 2-mg midazolam and administration of futibatinib 20 mg once daily, steady-state plasma concentrations of futibatinib were  AUC 0-inf , area under the plasma concentration-time curve from time 0 extrapolated to infinity; AUC 0-t , area under the plasma concentration-time curve from time 0 to time of last measured non-0 concentration; C max , maximum plasma concentration; CI, confidence interval; CL/F, oral clearance; %CV, percent coefficient of variation; GMR, geometric mean ratio; PK, pharmacokinetic; SD, standard deviation; t 1/2 , apparent first-order terminal elimination half-life; t max , time to reach maximum plasma concentration.
reached within 4 days of once-daily dosing as trough concentrations on days 4, 5, and 6 were comparable (Table S7).

Safety
No deaths, serious AEs, or discontinuations were reported in the 2 phase I DDI studies.In the futibatinib and itraconazole/rifampin study, 19 (48%) participants experienced AEs, most of which were grade 1 in severity.No grade 3 or 4 AEs were reported (Table S8).The most frequently reported AEs were diarrhea in part 1 (itraconazole) and headache in part 2 (ri-fampin).All AEs resolved by the end of study.Mean serum chemistry, hematology, and urinalysis parameters remained within normal limits at the assessed postdose time points.
In the futibatinib and midazolam study, 10 (42%) participants experienced AEs.All events were grade 1 except for 1 grade 2 event of abdominal pain (Table S9).The most frequently reported AEs were diarrhea (in 5 [21%] participants) and upper abdominal pain (in 4 [17%]).All AEs resolved without sequelae.Transient serum phosphate elevations (up to 2 times the upper limit of normal) were observed following daily futibatinib administration.Clinical laboratory, vital sign, and electrocardiogram data showed no clinically significant treatment-related trends.

Discussion
These results from in vitro and phase I studies provide a comprehensive overview of the interaction potential of futibatinib with CYP3A and P-gp.From in vitro CYP phenotyping studies, CYP3A4, CYP3A5, CYP2C9, and CYP2D6 were identified as potential CYP enzymes involved in futibatinib metabolism.Subsequent confirmatory studies showed CYP3A to be the major CYP isoform involved in futibatinib metabolism in the liver.In human liver S9 fraction assays, the contributions of CYP enzymes and GST to the overall metabolism of futibatinib were shown to be 41% and 6%, respectively.The remaining clearance of futibatinib (≈50%) was independent of CYP enzyme and GST activity and can likely be attributed to nonenzymatic glutathione conjugation. 10The contributions of CYP and non-CYP enzymes estimated here are consistent with other in vitro experiments involving human hepatocytes, 4 in which the contribution of CYP enzymes was estimated to be 40%-50%.In addition, the present in vitro analyses showed futibatinib to be a time-dependent inhibitor of CYP3A enzymes, and basic model analyses confirmed the clinical inhibition potential of futibatinib for CYP3A.Other in vitro experiments showed futibatinib did not induce CYP3A at clinically relevant concentrations (data on file).Overall, results of these in vitro studies indicated the need for clinical assessments of DDIs between futibatinib and perpetrators of CYP3A and sensitive CYP3A substrates.
In vitro drug transporter experiments showed that futibatinib is a P-gp substrate, suggesting the need for clinical assessments of DDIs between futibatinib and perpetrators of P-gp.The phase I studies described here assessed DDIs between futibatinib and dual perpetrators of CYP3A and P-gp (itraconazole and rifampin), and between futibatinib and midazolam (a sensitive CYP3A substrate).
Part 1 of the first phase I DDI study demonstrated that itraconazole 200 mg once daily coadministered with futibatinib 20 mg (single dose) markedly increased the plasma exposure of futibatinib (C max by 51% and AUC 0-t and AUC 0-inf by ≈41% each) relative to a single 20-mg dose of futibatinib administered alone in healthy participants.Based on this result, concomitant use of dual P-gp and strong CYP3A inhibitors was discouraged, as dose-limiting toxicities were observed in the 24-mg once-daily cohort of the phase I dose-escalation study and potential stronger CYP3A inhibitors such as ritonavir or cobicistat, which are also P-gp inhibitors, could cause increased futibatinib exposure. 2,27However, the <2-fold increase in plasma futibatinib exposure in the presence of itraconazole (a dual P-gp and strong CYP3A4 inhibitor) versus when administered alone suggests futibatinib was not a sensitive substrate of CYP3A4 or P-gp.Coadministration of futibatinib and itraconazole appeared to decrease the mean CL/F and mean t 1/2 of futibatinib relative to futibatinib administered alone; however, semilogarithmic plots of futibatinib concentration over time (Figure 2B) suggested that the mean elimination rate of futibatinib was not affected by itraconazole.To examine this result closely, the systemic elimination in individual participants was analyzed.Elimination was dramatically reduced in 3 of 20 participants (t 1/2 was 0.8-1.9,1.2-4.3,and 1.6-3.4hours in participants 10, 16, and 20, respectively; Figure S1), which suggested that these outliers could have disproportionally affected the results.To confirm the involvement of P-gp in futibatinib absorption and PK, we have planned an additional clinical DDI study using a P-gp-specific inhibitor (eg, quinidine).
In part 2 of the first phase I DDI study, the coadministration of a single 20-mg dose of futibatinib with the dual P-gp and strong CYP3A inducer rifampin (at 600 mg once daily) decreased futibatinib C max by 53% and AUC 0-t and AUC 0-inf by ≈64% each relative to a 20-mg dose of futibatinib administered alone.Given that the effect magnitude was slightly larger for AUC than C max and coadministration with rifampin was associated with reductions in futibatinib t 1/2 , without delaying t max , rifampin may affect not only efflux transport and the absorption process for futibatinib, but hepatic metabolism as well.
In the second phase I DDI study of futibatinib and midazolam, a sensitive CYP3A substrate, oral administration of 20-mg futibatinib once daily for 7 consecutive days showed no notable impact on midazolam exposure, indicating no significant DDIs between futibatinib and midazolam.Furthermore, no impact was noted on the exposure of total 1-OH-midazolam, which reflects in vivo CYP3A activity, based on similar metabolite to parent ratios for C max and AUC 0-inf observed with and without coadministered futibatinib, indicating that futibatinib likely does not affect the metabolic pathway of 1-OH-midazolam formation. 28The absence of in vivo DDI between futibatinib and midazolam suggests that the in vitro prediction model was not quantitatively translated to the in vivo setting.Overestimation of DDI in the basic model is not an unexpected observation, given the conservative approach, as previously reported. 26,29n the midazolam DDI study, 92% of the participants were men.A previous report found no sex differences in the magnitude of decreased apparent clearance of oral midazolam by the concomitant use of ketoconazole, a strong CYP3A inhibitor. 30Moreover, a population PK analysis revealed that sex was not a significant covariate of futibatinib PK (data on file).Taken together, these reports suggest that the DDI study results can be generalized to both male and female patients.The observed futibatinib exposure among healthy participants in this study (steady-state C max , 192.4 ng/mL; AUC, 835.5 ng • h/mL) was similar to that in patients with cancer. 2,5n both phase I DDI studies, futibatinib administered alone or coadministered with itraconazole, rifampin, or midazolam was generally safe and well tolerated in healthy adults, with most AEs being mild or moderate in severity.Overall, the safety profile of futibatinib, administered alone or coadministered with itraconazole, rifampin, or midazolam in healthy adults, was consistent with profiles observed previously in other phase I and II studies. 2,6,7n vitro experiments also showed inhibition of P-gp by futibatinib; unlike variations in CYP enzyme activity, the inhibition of P-gp is thought to have limited clinical importance, with the exception of alterations in the PK parameters of P-gp substrates with narrow therapeutic indices. 31To evaluate the potential risks of P-gpmediated DDIs, a physiologically based PK model was used to assess the impact of futibatinib on the plasma exposure of digoxin, a sensitive P-gp substrate.In this model, using the actual in vitro futibatinib IC 50 value for P-gp (assuming 20-mg futibatinib once-daily dosing), digoxin C max and AUC from time 0 to 96 hours increased by <10% (7% and 2%, respectively).When a 0.01-fold in vitro IC 50 value was used, digoxin C max and AUC from time 0 to 96 hours increased by 84% and 28%, respectively.A subsequent clinical study is planned to assess DDIs between futibatinib and sensitive P-gp substrates.

Conclusions
In line with in vitro data, results from phase I studies showed significant DDIs between futibatinib and itraconazole or rifampin, suggesting that concomitant uses of dual P-gp and strong CYP3A inhibitors and inducers with futibatinib should be avoided.In the midazolam study, futibatinib did not show inhibition of the CYP3A pathway in vivo, suggesting that futibatinib can be concomitantly used with other drugs metabolized by CYP3A.Given the in vitro inhibition potential of futibatinib for P-gp, additional DDI studies of futibatinib with P-gp sensitive substrates (eg, digoxin) and Pgp-specific inhibitors (eg, quinidine) are planned.Together, these data provide valuable guidance on using concomitant medications with futibatinib in patients with iCCA and other tumors and support further clinical investigation of futibatinib.

Figure 2 .
Figure 2. Mean plasma concentration-time profiles of futibatinib administered with or without itraconazole plotted in a linear (A) and semilogarithmic plot (B) or rifampin in linear (C) and semilogarithmic plots (D).Insets show expanded views of 0-8 hours.All values below the limit of quantitation are assumed to be 0. SD, standard deviation.
Figure 2C and D. Following administration of a single dose of futibatinib, mean futibatinib plasma concentrations increased, reaching a median t max by 1.3 hours (Figure 2C and D; Table

Figure 3 .
Figure 3. Mean plasma concentration-versus-time profiles of midazolam (A) and 1-OH-midazolam (B) following a single 2-mg dose of midazolam alone and in combination with futibatinib.Insets show expanded views of 0-8 hours.All values below the limit of quantitation are assumed to be 0. SD, standard deviation.

Table 1 .
Effect of Specific CYP Inhibitors on CL int of Futibatinib and Percentage of Contribution in Human Liver Microsomes.

Table 2 .
Concentration-Dependent Transcellular Transport of Futibatinib Across Human P-gp Expressing LLC-PK Cells and Mock-Transfected LLC-PK1 Cells.
P-gp, P-glycoprotein; SD, standard deviation.All directional transcellular transport assays were performed in triplicate; incubation time, 2 hours.

Table 3 .
26sessment of DDI Potential by Time-Dependent Inhibition of CYP3A.inact , maximal inactivation rate constant.R 2 is the predicted ratio of the victim drug's AUC in the presence and absence of futibatinib for basic models of time-dependent inhibition occurring in the liver and intestine, and was calculated according to the following equation26: R 2 = (k obs + k deg )/k deg , where k obs = (K inact × 50 × I max,u )/(K I,u + 50 × I max,u ).k obs is the observed (apparent first order) inactivation rate of the affected enzyme.

Table 4 .
Summary of Futibatinib PK Parameters Following a Single Dose of 20 mg of Futibatinib Alone (Reference) or in Combination with Itraconazole or Rifampin (Test) and Statistical Comparison.

Table 5 .
Summary of Midazolam PK Parameters Following a Single 2-mg Dose Midazolam Alone (Reference) or in Combination with Futibatinib (Test) and Statistical Comparison.