Clin Pharmacokinet DOI 10.1007/s40262-016-0461-9

REVIEW ARTICLE

Clinical Pharmacokinetics and Pharmacodynamics of Cabozantinib Steven A. Lacy1 • Dale R. Miles1,2 • Linh T. Nguyen1,3

Ó Springer International Publishing Switzerland 2016

Abstract Cabozantinib inhibits receptor tyrosine kinases involved in tumor angiogenesis and metastasis. The capsule formulation (CometriqÒ) is approved for the treatment of progressive metastatic medullary thyroid cancer at a 140-mg free base equivalent dose. The tablet formulation (CabometyxTM, 60-mg free base equivalent dose) is approved for the treatment of renal cell carcinoma following anti-angiogenic therapy. Cabozantinib displays a long terminal plasma half-life (*120 h) and accumulates *fivefold by day 15 following daily dosing based on area under the plasma concentration-time curve (AUC). Four identified inactive metabolites constitute [65 % of total cabozantinib-related AUC following a single 140-mg free base equivalent dose. Cabozantinib AUC was increased by 63–81 % or 7–30 % in subjects with mild/moderate hepatic or renal impairment, respectively; by 34–38 % with concomitant cytochrome P450 3A4 inhibitor ketoconazole;

Electronic supplementary material The online version of this article (doi:10.1007/s40262-016-0461-9) contains supplementary material, which is available to authorized users. & Steven A. Lacy [email protected] 1

Exelixis Inc., 210 East Grand Avenue, South San Francisco, CA 94080-0511, USA

2

Genentech Inc., South San Francisco, CA, USA

3

Medivation Inc., San Francisco, CA, USA

and by 57 % following a high-fat meal. Cabozantinib AUC was decreased by 76–77 % with concomitant cytochrome P450 3A4 inducer rifampin, and was unaffected following administration of proton pump inhibitor esomeprazole. Cabozantinib is a potent in vitro inhibitor of P-glycoprotein, and multidrug and toxin extrusion transporter 1 and 2-K, and is a substrate for multidrug resistance protein 2. No clinically significant covariates affecting cabozantinib pharmacokinetics were identified in a population pharmacokinetic analysis. Patients with medullary thyroid cancer with low model-predicted apparent clearance were more likely to dose hold/reduce cabozantinib early, and had a lower average dose through day 85. However, longitudinal tumor modeling suggests that cabozantinib dose reductions from 140 to 60 mg/day did not markedly reduce tumor growth inhibition in medullary thyroid cancer patients.

S. A. Lacy et al.

Key Points Cabozantinib, a receptor tyrosine kinase inhibitor whose targets include MET (hepatocyte growth factor receptor) and vascular endothelial growth factor receptor 2, is approved for the treatment of progressive metastatic medullary thyroid cancer at a 140-mg free base equivalent capsule dose and for the treatment of renal cell cancer following antiangiogenic therapy at a 60-mg free base equivalent tablet dose. Cabozantinib pharmacokinetics are characterized by a long terminal half-life (approximately 120 h), accumulation upon repeat daily dosing (approximately fivefold based on area under the plasma concentration-time curve [AUC]), and moderately high variability in exposure (38–43 % for AUC at steady state in cancer patients). Cabozantinib undergoes extensive metabolism: four major metabolites in human plasma comprise[65 % of total drug exposure (AUC), although each possesses minimal (\1/10th) on-target kinase inhibition potency relative to the parent drug. Cabozantinib plasma exposures may be increased by cytochrome P450 3A4 or multidrug resistance protein 2 inhibitors, a high-fat meal, or hepatic impairment. Cabozantinib may increase plasma concentrations of co-administered substrates of P-glycoprotein. Patients with medullary thyroid cancer with low model-predicted apparent clearance were more likely to dose hold/reduce cabozantinib. However, longitudinal tumor modeling suggests that cabozantinib dose reductions to 60 mg/day did not markedly reduce tumor growth inhibition in medullary thyroid cancer patients.

1 Introduction Cabozantinib inhibits multiple receptor tyrosine kinases implicated in tumor angiogenesis, invasion, or metastasis, including MET (hepatocyte growth factor receptor), vascular endothelial growth factor receptor 2 (VEGFR2), RET (the glial cell line-derived neurotrophic factor rearranged during transfection receptor), and AXL (the GAS6 receptor) [1]. Dysregulation of the MET/vascular endothelial growth factor (VEGF) axis is found in a number of human malignancies and has been associated with tumorigenesis.

VEGFR2 is expressed on endothelial cells, and is a wellestablished key mediator of VEGF signaling in the process of tumor angiogenesis [2]. MET is expressed on tumor cells and endothelial cells, and mediates hepatocyte growth factor signaling, leading to increased cell motility, proliferation, and survival [3]. MET upregulation occurs in a wide range of malignancies (e.g., thyroid, kidney, liver, prostate, ovarian, lung, and breast cancers), and is associated with more aggressive and invasive phenotypes of cancer cells in vitro and with metastases in vivo [4, 5]. MET-driven metastasis may be exacerbated by a number of factors, including tumor hypoxia caused by selective inhibition of the VEGF pathway. Clear cell renal cell carcinomas (RCCs) commonly exhibit mutations in the tumor suppressor von Hippel–Lindau gene [6], which triggers an increase in expression of VEGF and of MET, resulting in angiogenesis, tumor cell proliferation, and invasive growth [7, 8]. As cabozantinib inhibits both arms of the MET/VEGF axis, it may provide greater anti-tumor activity over targeting either pathway individually. In a double-blind phase III trial comparing the cabozantinib capsule formulation (CometriqÒ) at a daily dose of 140 mg free base equivalent (FBE) with placebo in 330 patients (randomized 2:1) with documented radiographic progression of metastatic medullary thyroid cancer (MTC), cabozantinib showed a statistically significant improvement for the primary study endpoint median progression-free survival (PFS) (hazard ratio [HR] 0.28; 95 % confidence interval [CI] 0.19–0.40; p \ 0.001) [9]. These results supported the approval of CometriqÒ in the USA for the treatment of progressive metastatic MTC, and in the European Union for the treatment of progressive, unresectable locally advanced or metastatic MTC. Commercially, CometriqÒ is available as 20- and 80-mg FBE hard-gelatin capsules in the European Union and in the USA. The recommended daily dose of CometriqÒ for the MTC indication is 140 mg FBE orally [10]. In an open-label phase III study, the cabozantinib tablet formulation (CabometyxTM) at a daily dose of 60 mg FBE was compared with everolimus at a daily dose of 10 mg in 658 patients (randomized 1:1) with RCC that had progressed after VEGFR-targeted therapy. A 60-mg FBE cabozantinib dose was evaluated in this phase III study, based on findings from a phase I study in patients with RCC of improved tolerance to study drug and evidence of clinical activity in patients who had dose reduced from 140 to 60 mg FBE [11]. Compared with the everolimus arm, cabozantinib showed a statistically significant improvement in both the primary study endpoint median PFS (HR 0.58; 95 % CI 0.45–0.75; p \ 0.001 [12]) and the secondary study endpoint overall survival (HR 0.66; 95 % CI, 0.53–0.83; p = 0.00026 [13]). These results supported the approval of CabometyxTM at a

Clinical Pharmacokinetics and Pharmacodynamics of Cabozantinib

60-mg FBE daily dose for the treatment of patients with RCC following anti-angiogenic therapy in the USA [14] or following prior VEGF-targeted therapy in the EU [15]. Cabozantinib tablets are also being evaluated in a pivotal clinical study in patients with hepatocellular cancer at a 60-mg FBE daily dose [16]. Dose reductions to a 40- or a 20-mg daily dose are permitted in both RCC and hepatocellular cancer studies to maintain treatment in response to drug-related adverse events. Commercially, CabometyxTM is supplied as yellow, film-coated, 20-, 40-, and 60-mg FBE tablets. Cabozantinib tablets use a more efficient and scalable manufacturing process considered preferable for higher volume commercial production than that used for capsule manufacture. Cabozantinib was characterized in a clinical pharmacology program that included pharmacokinetic (PK) studies in cancer patients following single or repeat dosing of cabozantinib capsules [17, 18], a bioequivalence (BE) study of the capsule and tablet formulations in healthy volunteers (HVs), and an evaluation of dose proportionality of the 20-, 40-, and 60-mg FBE cabozantinib tablets following single-dose administration in HVs [19]. Additional clinical pharmacology studies conducted with cabozantinib included evaluations of intrinsic factors (hepatic and renal impairment [20]) and extrinsic factors (a high-fat meal or a gastric pH-modifying agent [21]) on cabozantinib pharmacokinetics. Drug–drug interaction (DDI) studies evaluated the effects of cytochrome P450 (CYP)3A4 inducer rifampin and CYP3A4 inhibitor ketoconazole on cabozantinib pharmacokinetics in HVs, and the effects of steady-state cabozantinib plasma concentrations on the plasma pharmacokinetics of CYP2C8 substrate rosiglitazone in cancer patients [18]. The major metabolites of cabozantinib were structurally identified and their plasma PK profiles characterized in HVs following a single oral dose containing 14C-cabozantinib [22]. A population PK analysis was used to evaluate intrinsic/extrinsic factors for potential impact on cabozantinib pharmacokinetics in patients with MTC or glioblastoma multiforme. Individual model-predicted steady-state area under the plasma concentration-time curve (AUCss, pred), calculated under the assumption that the initial cabozantinib dose was maintained, was evaluated for potential correlation with time to first dose hold/modification, or average dose through day 85. The objective of these analyses was to determine if patients exhibiting lower cabozantinib apparent clearance (CL/F) were more likely to have a dose hold or reduction. Longitudinal tumor modeling was performed to investigate the impact of dose on tumor growth inhibition [23, 24].

2 Pharmacology and Pharmacodynamics 2.1 Preclinical Pharmacology Cabozantinib, a weak base, is weakly soluble in aqueous media, and demonstrates a pH-dependent solubility profile (0.11 mg/mL in 0.01 N HCl, and is practically insoluble at pH values C4). Cabozantinib also exhibits high cell permeability in vitro in MDCK cell assays (Papp values at 70–110 nm/s) [25]. Cabozantinib is thus considered to be a Biopharmaceutics Classification System Class II compound characterized by its low solubility and high permeability. Cabozantinib (S)-malate salt is available commercially in capsule (CometriqÒ) and tablet (CabometyxTM) formulations. The chemical structures of cabozantinib (XL184) free base and its major metabolites identified in human plasma are shown in Fig. 1. Cabozantinib is a multi-targeted inhibitor of receptor tyrosine kinases that functions by binding in a fully reversible manner to the kinase domain (including the ATP-binding site). In biochemical assays, cabozantinib was a potent inhibitor of MET, VEGFR2, RET, and AXL (IC50 values of 1.3, 0.035, 5.2, and 7 nM, respectively) [1]. Cabozantinib was also an effective inhibitor of MET-activating kinase domain mutations Y1248H, D1246N, and K1262R, with IC50 values of 3.8, 11.8, and 14.6 nM, respectively. In addition, parent cabozantinib showed approximately tenfold greater inhibition potency against kinase targets MET, RET, and VEGFR2 relative to EXEL5366, EXEL-5162, EXEL-1646, and EXEL-1644, its major metabolites present in human plasma [22]. In cellular assays, cabozantinib was also shown to inhibit phosphorylation of MET, VEGFR2, and AXL (IC50 values of 7.8, 1.9, and 42 nM, respectively) [1]. 2.2 Preclinical Pharmacodynamics Results from pharmacodynamic experiments have demonstrated that cabozantinib inhibits MET and VEGFR2 in multiple tumor models in vivo. Oral administration of cabozantinib resulted in blockade of MET phosphorylation in human lung tumor xenografts grown in nude mice, blockade of MET phosphorylation in livers of mice, and blockade of VEGFR2 phosphorylation in mouse lung tissue [1]. For both targets, the duration of action for cabozantinib was sustained, with [50 % inhibition observed for [8 h post-dose following a single 100-mg/kg dose. Oral administration of cabozantinib also resulted in blockade of phosphorylation of mutationally activated RET in human MTC xenografts in nude mice [26].

S. A. Lacy et al. Fig. 1 Cabozantinib and major metabolites EXEL-5162, EXEL-5366, EXEL-1644, and EXEL-1646 identified in human plasma

Treatment with cabozantinib caused potent anti-angiogenic effects in xenograft tumors (i.e., increased endothelial cell apoptosis and disruption of the vasculature occurring within 24 h post-dose), which resulted in significant tumor growth inhibition or regression in multiple tumor models, including MTC, RCC, hepatocellular cancer, and breast, lung, neuroblastoma, colorectal, and gastrointestinal stromal cancers [1, 26–32]. Cabozantinib blocked metastasis and improved overall survival in the transgenic RIP-Tag2 mouse model of pancreatic neuroendocrine cancer [33], and inhibited tumor growth in soft tissue and bone in human prostate cancer xenograft studies in mice [34]. Cabozantinib has also shown potent antitumor activity in sunitinib-resistant RCC mouse xenograft models with induced AXL and MET expression following long-term sunitinib exposure [27]. 2.3 Clinical Pharmacodynamics Serum levels of calcitonin and carcinoembryonic antigen were evaluated as potential biomarkers of tumor burden in

the pivotal phase III study in patients with MTC. Compared with baseline values, mean calcitonin and carcinoembryonic antigen levels at week 12 were significantly decreased (-45 and -23 %, respectively) in the cabozantinib treatment group relative to the placebo cohort (?57 and ?89 %, respectively; p \ 0.001) [9]. These changes in calcitonin and carcinoembryonic antigen levels from baseline to week 12 also showed a generally linear relationship with changes in target lesion size, suggesting these serum markers may be reflective of patient benefit. 2.4 Clinical Exposure-Response Analyses In the pivotal phase III study, 79 % of MTC patients (169 of 214) who received the 140-mg FBE cabozantinib dose eventually dose reduced [9]. Sixty-five percent (140 of 214) of these MTC patients dose reduced because of adverse events. Two protocol-defined cabozantinib dose reductions were allowed: from 140 to 100 mg/day, and from 100 to 60 mg/day. Adverse events were generally managed with concomitant medications or by dose

Clinical Pharmacokinetics and Pharmacodynamics of Cabozantinib

modification (interruptions or reductions); 85 % (183 of 214) of cabozantinib-treated subjects experienced dose modifications. Forty-two percent of subjects received 60 mg/day as their final dose [35]. Exposure response analyses showed an increased risk for time to first cabozantinib dose modification to be highly correlated with higher AUCss, pred (Fig. 2) [23]. However, Kaplan–Meier analyses of PFS stratified by time to first dose modification (tertiles) showed no clear association between early and late cabozantinib dose modifications and reduced PFS (Fig. 3). These findings suggest that, although

Fig. 2 Time to first dose modification Kaplan–Meier analysis stratified by AUCss,pred tertiles for cabozantinib-treated patients with medullary thyroid cancer. AUC area under the plasma concentrationtime curve, AUCss,pred model-predicted steady-state AUC

Fig. 3 Progression-free survival Kaplan–Meier analysis stratified by time to first dose modification tertiles for cabozantinib-treated patients with medullary thyroid cancer. FMOD time to first dose modification, PFS progression-free survival

MTC patients who clear cabozantinib more slowly may have a higher risk of early dose reduction relative to the faster clearing patients, the dose reduction paradigm used in the phase III study allowed individual MTC patients to ultimately attain a tolerated exposure providing clinical benefit. Non-linear mixed-effects models were developed to describe the relationship between cabozantinib exposure and target lesion tumor size in the phase III study in patients with MTC [24]. A piecewise, discontinuous, semiempirical tumor growth model adequately described both an early reduction phase (0–110 days) and a late stabilization phase (110–280 days) for patients on extended cabozantinib treatment. Emax models relating average model-predicted cabozantinib plasma concentration to average model-predicted change in tumor size predicted that average concentrations of 79 and 58 ng/mL, respectively, would provide 50 % of the maximum possible tumor reduction during the first 110 days of dosing, and during the subsequent 110–280 days of dosing. In comparison, higher mean steady-state plasma pre-dose concentrations of 1380 ng/mL were measured in MTC patients administered 140 mg cabozantinib FBE in this study (Table 1). Simulations of tumor responses showed that daily doses of 60 mg or greater were expected to provide a similar tumor reduction. Therefore, based on modeling of cabozantinib effects on target lesion size, the two protocol-defined cabozantinib dose reductions from 140 to 100 or from 100 to 60 mg/day were not projected to result in a marked reduction in cabozantinib efficacy. 2.5 Cardiac Electrophysiology No marked inhibition of human-ether-a-go-go (hERG) channel activity (IC50 values [30 lM) was observed for cabozantinib when tested in a manual patch-clamp assay in mammalian cell systems transfected with hERG cDNA [25]. In a separate assay, cabozantinib (10–30 lM) did produce a mild but significant hERG trafficking inhibition and a direct block of hERG channel activity (30 lM) in transfected mammalian cells. Metabolites EXEL-5162, EXEL-5366, EXEL-1646, and the non-sulfated form of EXEL-1644 were not hERG trafficking inhibitors or direct hERG channel blockers at concentrations tested up to 30 lM. The effect of orally administered CometriqÒ (140 mg FBE) on the QTc interval was evaluated in the phase III study in patients with MTC. A mean increase in QTcF of 10–15 ms was observed at 4 weeks after initiating CometriqÒ [10]. A concentration-QTc relationship could not be definitively established. Changes in cardiac wave form morphology or new rhythms were not observed. No CometriqÒ-treated patients had a QTcF [500 ms.

S. A. Lacy et al. Table 1 Single- and repeated-dose mean (%CV) PK parameters for a cabozantinib FBE capsule dose across studies in cancer patients Study type

Phase I (safety)a

Phase III (efficacy)a

Phase I (safety)a,f

Phase III (efficacy)a,f

Phase I (DDI with rosiglitazone)a

Subject population

Solid tumors

MTC

Solid tumors

MTC

RCC and DTC

Formulation

Capsule

Capsule

Capsule

Capsule

Capsule

Sampling

Dense

Sparse

Dense

Sparse

Dense

Dose, mg

140

140

140

140

C100e

Food intake

Fast 2 h before and 1 h after dose

Fast 2 h before and 1 h after dose

Fast 2 h before and 1 h after dose

Fast 2 h before and 1 h after dose

Fast 2 h before and 2 h after dose

Single dose

Repeated dose at steady state

Day

1

1

19

29

22

N

34–35

200

25–29

200

30–32

Cmax, ng/mL

570 (43)

541 (42)

2220 (37)

1640 (43)

1970 (39)

Tmax, hb

2 (2–24)

2.37 (1–6.62)

2 (0–25)

2 (0–6.67)

2 (0–25)

AUC0–24, hng/mL

8228 (34)

NC

37,850 (43)

NC

29,700 (38)

Predose conc, ng/mL

NC

NC

1710 (44)

1380 (53)

1484 (48)

Accumulation ratio

NA

NA

5.4 (64)c

3.6 (66)d

NC

AUC0–24 area under the plasma concentration-time curve from time zero to 24 h post-cabozantinib dose, Cmax maximum plasma concentration, conc concentration, CV coefficient of variation, DDI drug–drug interaction, DTC differentiated thyroid cancer, FBE freebase equivalent, MTC medullary thyroid cancer, NA not applicable, NC not calculated, PK pharmacokinetic, RCC renal cell carcinoma, Tmax time of the maximum concentration a

Data available at: http://www.accessdata.fda.gov/drugsatfda_docs/nda/2012/203756Orig1s000ClinPharmR.pdf

b

Data reported as median and range AUC ratio of day 19 to day 1

c d

Cmax ratio of day 29 to day 1

e

Data for subjects dosed at C100 (100 or 140) mg/day for C21 days prior to steady-state sampling

f

Subjects who were not at steady state were excluded from the summary statistics

3 Pharmacokinetic Characteristics 3.1 Comparison Between Single- and Multiple-Dose Pharmacokinetic Parameters The single-dose plasma pharmacokinetics for cabozantinib has been characterized in HVs administered the capsule and tablet formulations at FBE doses of 20–140 mg (Tables 2, 3, respectively). The median time to maximum plasma concentration (Tmax) values across these HV studies were from 3 to 5 h, although individual subjects did show prolonged absorption phases with maximum plasma concentration (Cmax) occurring as late as 120 h after dosing. Multiple peaks present in the plasma concentration–time profiles after a single oral dose suggest that cabozantinib is enterohepatically recirculated, or has delayed or multiple sites of absorption (Fig. 4). Although sulfate conjugate EXEL-1646 is the one human metabolite present at high plasma concentrations and detected in feces with a

structure capable of possible biotransformation back to the parent drug, it appears to be an unlikely source of enterohepatically recirculated cabozantinib. In vitro studies demonstrated that EXEL-1646 in the presence of sulfatase enzyme is converted back only to its non-conjugated monohydroxy precursor and not to the parent cabozantinib [22]. Following the absorption peak, plasma concentrations declined slowly with a mean terminal half-life of 111–131 h across the studies. PK parameter values at a single 140-mg FBE dose were generally consistent for capsule and tablet formulations administered in HVs and for the capsule formulation administered in subjects with advanced malignancies (Table 1). PK parameter values were also generally consistent across studies following repeat daily dosing of cabozantinib capsules at 100–140 mg FBE in patients with solid tumors, including MTC, RCC, and differentiated thyroid cancer. Higher cabozantinib exposure (Cmax and AUC from time zero to 24 h postcabozantinib dose [AUC0–24]) was observed after repeated

Clinical Pharmacokinetics and Pharmacodynamics of Cabozantinib Table 2 Single-dose mean (%CV) PK parameters of cabozantinib across studies after a 60- or 140-mg FBE capsule dose in HVs Study

Phase I (food effect)a,b

Phase I (DDI with rifampin)a,c

Phase I (DDI with ketoconazole)a,c

Phase I (bioequivalence)d

Phase I (hepatic impairment)a,e

Phase I (renal impairment)a,e

Population

HV

HV

HV

HV

HV

HV

Formulation

Capsule

Capsule

Capsule

Capsule

Capsule

Capsule

Dose (mg) Food intake

140 Fast 10 h before and 4 h after dose

140 Fast 10 h before and 4 h after dose

140 Fast 10 h before and 4 h after dose

140 Fast 10 h before and 4 h after dose

60 Fast 10 h before and 4 h after dose

60 Fast 10 h before and 4 h after dose

N

47

28

28

72

10

10

Cmax, ng/ mL

536 (38)

582 (45)

488 (41)

554 (43)

353 (21)

341 (28)

Tmax, hf

4 (2–24.03)

4 (1.98–24.08)

4 (1.13–24.05)

4 (2–5.04)

4 (2–5)

4 (3.00–4.03)

AUC0–t, hng/mL

59,200 (27)

55,500 (27)

47,600 (29)

54,900 (37)

31,100 (27)

29,720 (26)

AUC0–24, hng/mL

7420 (33)

7860 (38)

6220 (36)

6830 (35)

NR

3915 (20)

AUC0–inf, hng/mL

63,200 (28)

58,800 (28)

50,400 (32)

58,300 (39)

32,700 (29)

32,030 (27)

t‘, h

124 (24)

111 (27)

122 (33)

112 (26)

108 (26)

126 (22)

AUC0–24 area under the plasma concentration-time curve from time zero to 24 h post-cabozantinib dose, AUC0–t area under the plasma concentration-time curve from time zero to the time of the last measurable concentration, AUC0–inf area under the plasma concentration-time curve from time zero to infinity, BE bioequivalence, Cmax maximum observed concentration, CV coefficient of variation, DDI, drug–drug interaction, FBE freebase equivalent, HV healthy volunteer, NR not reported, PK pharmacokinetic, Tmax time of the maximum concentration, t1/2 apparent terminal elimination half-life a

Data are for the reference (no interacting treatment) or matched healthy control group

b

Nguyen et al. [21]

c

Nguyen et al. [18]

d

Nguyen et al. [19]

e

Nguyen et al. [20]

f

Median (range) was reported for Tmax

daily dosing, with mean accumulation ratios of 5.4-fold (AUC) and 3.6-fold (Cmax). Repeated daily dosing of cabozantinib has not been conducted in HVs. 3.2 Comparison of Cabozantinib Pharmacokinetics for Capsule and Tablet Formulations A BE study of cabozantinib capsule and tablet formulations was conducted in HVs receiving a single 140-mg FBE dose [19]. PK parameter values for this study are presented in Tables 2 and 3; PK plasma concentration over time plots for both formulations are presented in Fig. 4. Mean PK parameters were moderately variable with percent coefficients of variation (%CVs) for the main parameters ranging from 44 to 47 % (AUC) and 54 % (Cmax) for cabozantinib tablets and from 35 to 39 % (AUC) and 43 % (Cmax) for cabozantinib capsules. The median Tmax was 3.5 h for the tablet formulation and 4.0 h for the capsule formulation; the terminal half-life of cabozantinib appeared to be similar for both formulations (mean values of 112–115 h).

In the statistical BE evaluation of cabozantinib capsules and tablets, the geometric least squared (GLS) mean values for AUC0–t and AUC from time zero to infinity (AUC0–inf) values were slightly (8 %) higher for the tablet formulation treatment relative to the capsule formulation treatment. The 90 % CIs around the ratios of GLS mean values were within the pre-specified bioequivalence limits of 80.00–125.00 % for both AUC0–t and AUC0–-inf. The Cmax value was 19 % higher for the tablet formulation cohort compared with the capsule formulation cohort, and the upper limit of the 90 % CI around the ratio percent of GLS mean value (131.65 %) was outside the standard BE limit acceptance range of 80.00–125.00 % [36–38]. Therefore, the BE of the two formulations was not demonstrated. 3.3 Characteristics of Drug Absorption The absolute bioavailability of cabozantinib tablet or capsule formulations has not been determined. In a phase I dose escalation study in cancer patients [17], the capsule

S. A. Lacy et al. Table 3 Single-dose mean (%CV) PK parameters for cabozantinib tablet doses of 20, 40, 60, 80, or 140 mg FBE in HVs Study

Phase I (bioequivalence)b

Phase I (DDI with esomeprazole)c,d

Phase I (pharmacokinetics)b

Population

HV

HV

HV

Formulation

Tablet

Tablet

Tablet

Tablet

Tablet

Dose (mg)

140

80

60

40

20

HV

HV

Food intake

Fast 10 h before and 4 h after dose

Fast 10 h before and 4 h after dose

Fast 10 h before and 4 h after dose

N

72

21

21

21

21

Cmax, ng/mL

702 (54)

647 (30)

343 (41)

239 (56)

117 (72)

Tmax, h AUC0–t, hng/mL

3 (2–24) 61,900 (44)

3 (2–5) 55,800 (25)

4 (2–8) 29,800 (38)

3 (2–48) 19,800 (42)

3 (1–120) 9290 (50)

AUC0–24, hng/mL

8140 (47)

7580 (31)

3880 (33)

2620 (53)

1280 (59)

AUC0–inf, hng/mL

65,800 (46)

58,900 (25)

32,100 (39)

21,100 (42)

10,400 (48)

AUC0–inf /dose, hng/mL/ mg

470

736

535

528

520

t‘, h

115 (31)

117 (25)

111 (18)

122 (22)

131 (25)

a

AUC0–24 area under the plasma concentration-time curve from time zero to 24 h post-cabozantinib dose, AUC0–t area under the plasm concentration-time curve from time zero to the time of the last measurable concentration, AUC0–inf area under the plasma concentration-time curve from time zero to infinity, BE bioequivalence, Cmax maximum observed concentration, CV coefficient of variation, DDI drug–drug interaction, FBE freebase equivalent, HV healthy volunteer, PK pharmacokinetic, Tmax time of the maximum concentration, t1/2 apparent terminal elimination half-life a

Median (range) was reported for Tmax

b

Nguyen et al. [19]

c

Nguyen et al. [21]

d

Data are for the reference (no interacting treatment) group

1200

Mean (SD) Cabozantinib Concentration (ng/mL)

1000 800

1200 600 400

1000

200

800

0 0

6

12

18

24

Time Postdose (hr)

600

Treatment A, Tablet (n=72) Treatment B, Capsule (n=72)

400

study [22], the cabozantinib formulation (140 mg FBE containing 100 lCi [14C]-cabozantinib) was administered as a solution. Mean AUC0–inf values for HVs administered a 140-mg cabozantinib capsule dose [1, 21] or tablet dose [18] were 74–93 % and 97 %, respectively, of the corresponding value in the mass balance study where cabozantinib was formulated as a solution. This suggests a potentially high bioavailability of the capsule and tablet formulations. 3.4 Characteristics of Drug Distribution

200 0 0

72

144

216

288

360

432

504

Time Postdose (hr)

Fig. 4 Mean (±SD) cabozantinib plasma concentrations over time plots following a single oral 140-mg dose of cabozantinib capsule or tablet formulation in healthy volunteers (n = 72) 0–504 h after dose; inset 0–24 h after dose. SD standard deviation

formulation yielded approximately twofold higher dosenormalized AUC0–24 h after a single dose compared with an oral liquid suspension formulation. In the mass balance

Plasma protein binding of cabozantinib was evaluated by equilibrium dialysis in vitro (pre-dose samples fortified with cabozantinib) and in vivo (collected 4 h after dosing with cabozantinib) in subjects with hepatic impairment [20, 25]. In the in vitro assay, the mean protein binding of cabozantinib (2 lM initial concentration) was approximately equivalent in the mild Child–Pugh (CP)-A and moderate CP-B hepatic impairment groups and matched HVs (99.79, 99.77, and 99.76 %, respectively). In the 4-h post-dose in vivo samples, the mean protein binding of cabozantinib in the CP-A and HV control groups was

Clinical Pharmacokinetics and Pharmacodynamics of Cabozantinib

similar (approximately 99.67 and 99.65 %, respectively) and was slightly lower in the CP-B group (approximately 99.43 %). The percentage of cabozantinib unbound to plasma protein appeared to be correlated with serum albumin concentration; subjects with low serum albumin concentrations exhibited a higher percentage of unbound plasma protein and lower total Cmax. The protein binding of the major plasma metabolite 6-desmethyl amide cleavage product sulfate (EXEL-1644) in human plasma was determined in vitro by equilibrium dialysis at concentrations of 125–500 lM [21]. The mean percentages of EXEL-1644 bound to human plasma proteins were very high (range: 99.950–99.996 %). The concentration of total radioactivity in the plasma and the concentration of total radioactivity in the blood were measured in the mass balance study following a single 140-mg cabozantinib FBE dose containing 14Ccabozantinib [22]. The mean values of systemic exposures (AUC0–24 and AUC0–72) in plasma were around 1.6 times higher than those in whole blood. The mean percent total radioactivity concentration present in erythrocytes relative to whole blood within 72 h post-dose (range: 0.174 ± 4.51 % to 12.3 ± 3.71 %) indicated that radioactivity was present primarily in plasma and not markedly associated with red blood cells. 3.5 Evidence of Hepatic or Renal Routes of Elimination In the mass balance study in HVs [22], 27.29 ± 4.65% and 53.79 ± 4.52% (mean ± standard deviation) of the administered radioactive dose (140 mg cabozantinib FBE containing 100 lCi [14C]-cabozantinib) was recovered in the urine and feces, respectively. A mean (±standard deviation) percent recovery of 81.09 ± 1.56 % (range: 78.14–83.38 %) of the total radioactivity dose was recovered in combined urine and feces through 48 days postdose. Approximately 1 % total mean radioactivity was recovered in combined feces and urine after day 28 postdose. Parent cabozantinib and metabolites were measured by radio-quantitation in pooled human fecal and urine samples from six HVs in the mass balance study. Percent parent cabozantinib-associated radioactivity relative to total fecal radioactivity in individual time pooled fecal samples from six individual subjects ranged from 22 to 67 %, below quantifiable limits-42 %, and below quantifiable limits-44 % from 0 to 72 h, 144–192 h, and 288–336 h post-dose, respectively. Parent cabozantinib was not observed in urine samples. Thus, hepatobiliary elimination appears to be the major route of elimination of cabozantinib and its metabolites, while urinary excretion is a route of elimination exclusively for cabozantinib metabolites.

3.6 Characteristics of Cabozantinib Metabolism The major metabolites of cabozantinib (XL184) identified by LC-MS/MS methods in human plasma following a single oral dose (140 mg cabozantinib FBE containing 100 Ci [14C]-cabozantinib [22]) in HVs were XL184-amide cleavage product [EXEL-5366], XL184-N-oxide [EXEL5162], XL184-monohydroxy sulfate [EXEL-1646], and 6-desmethyl amide cleavage product sulfate [EXEL-1644]. For cabozantinib and its four major metabolites EXEL5366, EXEL-5162, EXEL-1646, and EXEL-1644, mean exposure ratios relative to total exposure [AUC0–t (each analyte)/AUC0–t (parent ? 4 measured metabolites)] were 32.0, 3.09, 4.90, 13.8, and 45.9 %, respectively. Metabolite 4-fluoroaniline concentrations were below quantifiable limits for all subjects. Median plasma Tmax occurred at approximately 1.49, 18.99, 13.50, 24.00, and 168 h with corresponding mean Cmax values of 1250, 52.9, 118, 236, and 230 ng/mL, respectively. The mean plasma elimination half-life values for cabozantinib, EXEL-5366, EXEL-5162, and EXEL-1646 were estimated to be 102, 91.8, 89.2, and 86.0 h, respectively. The plasma-concentration profile for metabolite EXEL-1644 shows slow accumulation over time with minimal change in plasma concentrations from 3 to 28 days following single-dose cabozantinib administration. The low plasma clearance of EXEL-1644 may reflect its high plasma protein-binding affinity. A plasma half-life value for EXEL-1644 could not be accurately calculated based on incomplete characterization of the terminal elimination phase. However, based on its long estimated half-life (*500 h), EXEL-1644 would be the predominant metabolite in plasma following repeat daily dosing of cabozantinib, with steady plasma exposures comparable to or higher than the parent drug. 3.7 Evidence of Pharmacokinetic Parameter Dose Linearity The relationship between three different cabozantinib tablet strengths (20, 40, and 60 mg FBE) and their respective PK parameter values was assessed in a phase I study in HVs [19]. The 20- and 60-mg tablet strengths were the cabozantinib formulations evaluated in the pivotal phase III study in RCC. Single doses of cabozantinib tablets at 20-, 40-, and 60-mg FBE dose strengths showed dose-proportional increases in mean plasma cabozantinib Cmax and AUC0–t values (Fig. 5). The 95 % CIs for the slopes of the ln-transformed Cmax, AUC0–t, and AUC0–inf vs. ln-transformed cabozantinib dose following single-dose administration of 20-, 40-, and 60-mg FBE cabozantinib tablet strengths (0.7849–1.6719, 0.8410–1.4658, and 0.7818–1.4353, respectively) included the value of 1. Therefore, dose proportionality was concluded for the

S. A. Lacy et al.

state (Table 1). Exposure variability for cabozantinib in cancer patients and HVs appears similar.

500

400

Plasma Cabozantinib Concentration (ng/mL)

300

4 Effect of Intrinsic Factors on Cabozantinib Pharmacokinetics

200

500

100

400

0 0

6

12

18

24

4.1 Effects of Sex, Body Mass, Age, and Race (Population Pharmacokinetic Analysis)

300

200 Treatment A: 20 mg Treatment B: 40 mg Treatment C: 60 mg

100

0 0

72

144

216

288

360

432

504

Time from Dosing (hr)

Fig. 5 Mean (±SD) cabozantinib plasma concentrations over time plots following a single oral dose of a 20-, 40-, or 60-mg cabozantinib tablet in healthy subjects (n = 21), 0–504 h after dose; inset 0–24 h after dose. SD standard deviation

single-dose PK parameters Cmax, AUC0–t, and AUC0–inf. CL/F was also similar across tablet strengths. Dose-normalized exposures (AUC0–inf/dose) for the 20-, 40-, and 60-mg FBE cabozantinib tablet strengths are similar to that of the single 140-mg FBE cabozantinib tablet dose (1 9 100 mg ? 2 9 20 mg) evaluated in the BE study, suggesting dose proportionality over the 20- to 140-mg tablet dose range (Table 3). The single-dose exposure (AUC) at the 60-mg tablet dose also appears to be comparable to single-dose exposures at the 60-mg capsule dose in healthy subject cohorts (Table 2). This finding further supports the conclusion that cabozantinib capsules and tablets yield similar plasma PK exposures when administered over a broad range of clinically relevant doses. 3.8 Inter- and Intra-Subject Variability of Pharmacokinetic Parameters The inter-subject variability (%CV) in HVs following a single capsule or tablet dose ranged from 20 to 59 % for AUC values and from 28 to 72 % for Cmax across the studies (Tables 2, 3). The within-subject variability (%CV) was estimated to be 39 % for Cmax and 28 % for AUC values in the capsule-tablet bioequivalence study. The inter-subject variability in cancer patients (%CV) was 42–43 % for Cmax and 34 % for AUC after a single dose, and 37–43 % for Cmax and 38–43 % for AUC at steady

A population PK analysis was performed on pooled data for 289 cabozantinib-treated cancer patients (including MTC) receiving daily administration of the cabozantinib capsule formulation at a dose of 140 mg FBE/day, except for five subjects that were dosed at 200 mg FBE/day [23]. The data were adequately described by a one-compartment model with first-order absorption and first-order elimination with a small lag time. The mean CL/F and apparent volume of distribution (Vc/F) were estimated to be 106 (standard error [SE] %: 2.98 %) L/day and 349 (SE %: 2.73 %) L, respectively, resulting in an estimated effective halflife of 2.28 days (55 h). Inter-individual variability in CL/F was modest (CV approximately 35 %). Sex and body mass index (BMI) on oral clearance were the only covariates retained in the final model. Cabozantinib CL/F was shown to decrease approximately 22 % at the 95th percentile BMI relative to the median, which translates to an approximately 28 % increase in steady-state AUC. Female individuals cleared cabozantinib more slowly than male individuals by 22 %, resulting in a predicted increase in steady-state AUC of 28 % for female individuals. Sex and BMI combined contributed 15 % to the variability in cabozantinib CL/F, and thus individually were not considered to be clinically meaningful to warrant dose adjustment. At the 95 % percentile BMI in female individuals, CL/F decreased approximately 41 % relative to the population estimate. Although this population would constitute a small percentage of patients receiving cabozantinib, dose reduction should be considered upon identification of treatment-emergent adverse events that are possibly related to comparatively higher cabozantinib exposure. Cabozantinib pharmacokinetics was not affected by age (20–86 years). No effect of race on cabozantinib pharmacokinetics could be concluded based on the low percentages of non-white patients (\4 % for individual races; \9 % for combined races) in the study population. 4.2 Hepatic and Renal Impairment Two clinical pharmacology studies were conducted to characterize single-dose pharmacokinetics of cabozantinib in renal- or hepatic-impaired subjects [20]. All subjects

Clinical Pharmacokinetics and Pharmacodynamics of Cabozantinib

received one 60-mg cabozantinib oral capsule dose followed by PK sampling over 21 days. The renal impairment study enrolled ten subjects each with mild or moderate impairment of renal function; 12 HVs were matched to the moderate group for age, sex, and BMI. The GLS mean ratios (as percentages of reference) for plasma cabozantinib AUC0–inf for impaired to normal organ function cohorts were approximately 30 % (GLS mean: 130.1; 90 % CI 98.8–171.3) and 7 % (GLS mean: 106.6; 90 % CI 79.6–140.1) higher in subjects with mild and moderate renal impairment, respectively. Based on the small exposure increase, cabozantinib should be used with caution in patients with mild or moderate renal impairment with no recommended adjustment in the starting dose. Cabozantinib has not been evaluated in subjects with severe renal impairment. The hepatic impairment study enrolled eight male individuals each with mild or moderate hepatic impairment; ten healthy male individuals were matched to the moderate group for age, BMI, and ethnicity. The GLS mean ratios (as percentages of reference) for plasma cabozantinib AUC0–inf for impaired to normal organ function cohorts were approximately 81 % (GLS mean ratio: 181.2; 90 % CI 121.4–270.3) and 63 % (GLS mean ratio: 162.7; 90 % CI 107.4–246.7) higher in subjects with mild and moderate hepatic impairment, respectively. A reduced cabozantinib starting dose is recommended in patients with mild or moderate hepatic impairment [10, 14, 15]. Cabozantinib has not been evaluated in subjects with severe hepatic impairment. 4.3 Pediatric Use The pharmacokinetics of cabozantinib has not yet been adequately characterized in the pediatric population.

5 Effect of Extrinsic Factors on Cabozantinib Pharmacokinetics Extrinsic factors evaluated in vitro and in vivo that may affect cabozantinib clinical pharmacokinetics include possible DDIs affecting CYP-mediated cabozantinib metabolism and the effects of food intake and gastric pH modification. Although not extrinsic factors, the DDI potential of cabozantinib as a CYP inhibitor or inducer has also been evaluated. The effects on cabozantinib PK by CYP activity modifiers, food intake, and gastric pH modification observed in the clinical pharmacology studies are presented in Supplemental Fig. 1. Examinations of cabozantinib as a substrate and/or inhibitor of clinically relevant drug transporters were conducted in in vitro assays only.

5.1 Potential DDIs: Cabozantinib as a Cytochrome P450 Substrate Cabozantinib (XL184) was shown to be a substrate for CYP3A4 metabolism in vitro, as a neutralizing antibody to CYP3A4 inhibited formation of a cabozantinib-derived metabolite (XL184 N-oxide; EXEL-5162) by [80 % in a NADPH-catalyzed, human liver microsomal incubation [25]. In contrast, neutralizing antibodies to CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C19, CYP2D6, and CYP2E1 had no effect on cabozantinib metabolite formation. A neutralizing antibody to CYP2C9 showed a minimal effect on cabozantinib metabolite formation (i.e., a \20 % reduction). The effects of CYP3A4 inhibitor ketoconazole and CYP3A4 inducer rifampin on the plasma pharmacokinetics of cabozantinib were investigated in two separate clinical pharmacology studies [18]. A crossover single-sequence study compared cabozantinib plasma pharmacokinetics in 28 HVs following a single 140-mg cabozantinib FBE dose alone or during ketoconazole administration at steady-state plasma levels (day 7 at 400 mg daily dosing for 27 days). The GLS mean values for Cmax, AUC0–t, and AUC0–inf for cabozantinib following co-administration of ketoconazole decreased by 3 %, and increased by 34 % and 38 %, respectively, compared with the values when cabozantinib was administered alone. The GLS mean ratios (as percentages of reference) were: 97.37 % (90 % CI 83.07–114.11) for Cmax, 134.30 % (90 % CI 122.45–147.30) for AUC0–t, and 138.05 (90 % CI 124.51–153.07 %) for AUC0–inf. The findings from this study indicate that the CYP3A4 substrate cabozantinib is susceptible to interaction with ketoconazole, a potent CYP3A4 inhibitor. Therefore, concomitant use of strong inhibitors of CYP3A4 should be avoided as they would be anticipated to potentially increase systemic exposure (AUC values) of cabozantinib. A reduced cabozantinib starting dose is recommended in patients receiving long-term treatment of a strong CYP3A4 inhibitor [10]. The effect of the CYP3A4 inducer rifampin was investigated in a crossover single-sequence study comparing cabozantinib plasma pharmacokinetics in 28 HVs following a single cabozantinib 140-mg FBE dose alone or with concomitant rifampin administration at steady state (day 11 of 600 mg daily for 31 days). The GLS mean values for Cmax, AUC0–t, and AUC0–inf for cabozantinib following coadministration of rifampin increased by 8 %, and decreased by 76 and 77 %, respectively, compared with cabozantinib administered alone. The GLS mean ratios (as percentages of reference) for Cmax, AUC0–t, and AUC0–inf of cabozantinib following co-administration with rifampin compared with cabozantinib administered alone were: 107.84 % (90 % CI 94.38–123.23) for Cmax, 24.25 % (90 %

S. A. Lacy et al.

CI 22.11–26.59) for AUC0–t, and 23.03 % (90 % CI 20.89–25.40) for AUC0–inf. Based on the marked decrease in cabozantinib exposure (AUC) with potent CYP3A4 inducer rifampin, long-term concomitant use of strong inducers of CYP3A4 should be avoided as they would be anticipated to potentially decrease systemic exposure (AUC values) of cabozantinib. An increased cabozantinib starting dose is recommended in patients receiving longterm treatment of a strong CYP3A4 inducer [10]. 5.2 Potential Drug–Drug Interactions: Cabozantinib as a Cytochrome P450 Inhibitor In in vitro CYP inhibition studies, CYP2C8 was the isozyme most potently inhibited by both cabozantinib and the predominant circulating metabolite EXEL-1644 (Ki = 4.6 and 1.1 lM, respectively), followed by CYP2C9 (Ki = 10.4 and 32.5 lM, respectively) [22]. Cabozantinib was a weaker direct inhibitor of isozymes CYP2B6, CYP2C19, and CYP3A4/5 (midazolam 10 -hydroxylase) with IC50 values of 10.1, 40, and 272 lM, respectively. IC50 values of [100 lM were determined for cabozantinib for CYP1A2, CYP2D6, or CYP3A4/5 (testosterone 6b-hydroxylase), and for EXEL-1644 for CYP1A2, CYP2B6, CYP2C19, CYP2D6, and CYP3A4/5 [both midazolam 10 -hydroxylase and testosterone 6b-hydroxylase]). Neither cabozantinib nor EXEL-1644 was shown to be a metabolism-dependent inhibitor of CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, or CYP3A4/5. As CYP2C8 was the isozyme most potently inhibited by cabozantinib in vitro, a phase I, single-arm, open-label, two treatment, two-period, single-sequence crossover study in cancer patients (n = 40) was conducted to determine if clinically relevant steady-state cabozantinib exposures (C100 mg FBE/day for C21days) affected the plasma pharmacokinetics of a single dose of rosiglitazone (4 mg), a CYP2C8 substrate drug [18]. No significant difference was observed in rosiglitazone plasma Cmax, AUC0–24, or AUC0–-inf values at steady-state cabozantinib plasma concentrations on day 22 compared with day 1 prior to initiation of cabozantinib treatment. The GLS day22:day1 mean ratios (90 % CI) for rosiglitazone Cmax, AUC0–24, and AUC0–inf values were 1.0396 (0.9261–1.1671), 1.0464 (0.9906–1.1053), and 1.0656 (1.0080–1.1265), respectively; these values fell within the standard 90 % CI bioequivalence acceptance limits of 0.8 and 1.25. Based on these data, plasma concentrations of cabozantinib and metabolite EXEL-1644 resulting from daily cabozantinib administration (C100 mg FBE/day for C21days) are not anticipated to result in any clinically relevant DDIs resulting from inhibition of CYP2C8. By inference, clinically relevant exposures of cabozantinib/ EXEL-1644 are also unlikely to impact other CYP

isozymes less potently inhibited by cabozantinib/EXEL1644 in vitro. A clinical DDI (increased coagulation times) reported in a RCC patient receiving concomitant cabozantinib and warfarin was attributed to cabozantinibrelated inhibition of CYP-mediated warfarin metabolism [39]; however, this event may reflect displacement of plasma protein-bound warfarin by cabozantinib. 5.3 Potential DDIs: Cabozantinib as a Cytochrome P450 Inducer Cabozantinib (30 lM) exposure resulted in a marked increase in CYP1A1 mRNA levels in primary human hepatocyte cultures (i.e., a [20-fold induction relative to negative controls and 75–100 % of the magnitude of effect of positive control b-naphthoflavone) [25]. Cabozantinib (B10 lM) exposure also resulted in a sevenfold maximal increase in CYP3A4 mRNA levels (i.e., approximately 90 % of the magnitude of positive control rifampicin), whereas CYP2C8, CYP29, and CYP2C19 mRNA levels were largely unaffected (B2-fold increase). In a separate assay using primary hepatocyte cultures, cabozantinib exposure (B20 lM) resulted in no biologically significant increases (B40 % of adjusted positive control response; [40, 41]) in CYP1A2-, CYP2B6-, or CYP3A4-mRNA levels or associated enzyme activities. These in vitro study findings suggest cabozantinib maybe a potential inducer of CYP1A1 in vivo; however, this isozyme is generally not a major biotransformation pathway for many concomitant medications. Cabozantinib exposure resulted in demonstrable CYP3A4 mRNA induction in freshly isolated human hepatocytes in one of two independent in vitro assays. Although CYP3A4 induction potential has not been specifically evaluated in a clinical pharmacology study, no marked effect on the plasma pharmacokinetics of the CYP3A4 substrate erlotinib was observed when co-administered with cabozantinib in a phase I study in patients with solid tumors (unpublished finding). Cabozantinib has little or no in vivo induction potential for isozymes CYP1A2, CYP2B6, CYP2C8, CYP2C9, or CYP2C19 based on in vitro study results. 5.4 Effects of Food Intake and Gastric pH Modification on Cabozantinib Pharmacokinetics In a phase I food-effect study of cabozantinib, male and female HVs received a single 140-mg FBE cabozantinib capsule dose under fasting and fed (high-fat breakfast) conditions, with a minimum 28-day washout between periods [21]. The GLS fed:fasted mean ratio percent values (90 % CIs) were 140.51 % (117.93–167.41) for Cmax, 157.37 % (135.75–182.44) for AUC0–t, and 156.9 %

Clinical Pharmacokinetics and Pharmacodynamics of Cabozantinib

(135.13–182.31) for AUC0–inf. The half-life values appeared to be similar following treatment under fed and fasted conditions, suggesting food ingestion had no effect on cabozantinib elimination processes. The median Tmax was longer in the presence of food (i.e., increased from 4 [fasted] to 6 [fed] h). As a high-fat meal was shown to significantly increase cabozantinib systemic exposure in this study, administration of cabozantinib is recommended in the fasted state. Although this food-effect study used cabozantinib capsules and the results may be formulation dependent, standard conservative dosing guidance is proposed (i.e., food fasting 2 h before and 1 h after cabozantinib administration) for both CometriqÒ and CabometyxTM to eliminate the risk of any clinically meaningful effects of food intake on plasma exposure. The effect of daily administration of proton pump inhibitor esomeprazole on the single-dose plasma PK of cabozantinib was evaluated in a phase I study in HVs [21]. The GLS mean values for Cmax, AUC0–t, and AUC0–inf were similar (between approximately 108 and 111 %) for a single 100-mg FBE cabozantinib dose administered alone or following six daily doses of 40 mg esomeprazole. The 90 % CIs around the ratio of GLS means were within the limits of 80.00–125.00 % for AUC0–t and AUC0–inf parameters; however, the upper 90 % CI for Cmax was determined to be 125.1 %. Co-administration of multiple doses of esomeprazole with a single dose of cabozantinib did not result in a clinically meaningful change in cabozantinib plasma exposure. Therefore, concomitant dosing of cabozantinib with proton pump inhibitors or weaker gastric pH-altering agents is permissible owing to the low risk of a clinically significant DDI. 5.5 Cabozantinib as a Substrate and/or Inhibitor of Drug Transporters Cabozantinib and EXEL-1644 were evaluated in vitro as potential substrates or/or inhibitors of clinically relevant drug transporters [22]. Cabozantinib was shown to be a substrate for transporter multidrug resistance protein (MRP)2 only, whereas EXEL-1644 appears to be a substrate for several drug transporters (i.e., organic anion transporter [OAT] 3, organic anion transporting polypeptide [OATP]1B1, OATP1B3, breast cancer resistance protein [BCRP], and MRP2). Assay data were inconclusive as to whether EXEL-1644 is a P-glycoprotein (P-gp) substrate. Cabozantinib demonstrated inhibition of multidrug and toxin extrusion (MATE) transporter 1 (MATE1) and 2-K (MATE2-K) (estimated IC50 values of 5.94 and 3.12 lM, respectively), but no marked inhibition of BCRP, bile salt export pump (BSEP), and MRP2 (i.e., IC50 values [50 lM), or of organic cation transporter 1 (OCT1), OAT1, OAT3, OATP1B1, and OATP1B3 (i.e., IC50 values [15

lM). Cabozantinib inhibited P-gp in vitro, with a lower IC50 value determined in MDCK-MDR1 cells (7 lM [25]) than in Caco-2 cells ([50 lM [22]). EXEL-1644 demonstrated the most potent inhibition of OAT1, OAT3, and OATP1B1 (IC50 range: 4.3–6.1 lM), less potent inhibition of BSEP, MRP2, OATP1B3, MATE1, and MATE2-K (IC50 range: 16.7–78.5 lM), and no marked inhibition of BCRP, P-gp, OCT1, and OCT2 (i.e., IC50 values exceeded the assay incubation solubility limit of 250 lM). Cabozantinib and EXEL-1644 at clinically relevant plasma exposures may represent potential risk of a drug transporter DDI. As both were shown to be drug transporter substrates, their plasma pharmacokinetics may also be affected by drugs that inhibit these transporters (e.g., administration of MRP2 inhibitors with cabozantinib should be approached with caution [10, 14, 15]). In general, the drug transporter IC50 values determined for cabozantinib and EXEL-1644 suggest a low potential risk for inhibiting drug transporters at clinically relevant steady-state plasma concentrations. US Food and Drug Administration and European Medicines Agency decision tree criteria for conducting in vivo DDI studies for most drug transporters are based on estimated unbound (freefraction) steady-state plasma drug concentrations [40, 41]. As both cabozantinib and EXEL-1644 have high plasma protein-binding affinities that yield low estimated steadystate free-fraction concentrations, regulatory guidance thresholds for conducting drug transporter inhibition in vivo DDI studies were generally not met. One exception is the US Food and Drug Administration decision tree criteria for transporter P-gp that uses a total (bound and free) steady-state drug Cmax value in its assessment of whether an in vivo DDI study should be conducted. As the ratio is [0.1 for a total steady-state cabozantinib plasma concentration value at clinically relevant doses (approximately 4.4 lM at 140 mg FBE daily) relative to the lowest cabozantinib P-gp IC50 value (7 lM), cabozantinib appears to have the potential to inhibit P-gp activity in vivo. Cabozantinib-mediated inhibition of P-gp efflux transporter activity has been implicated in the enhanced cytotoxicity of P-gp substrate anticancer drugs in human hepatoma HepG2 cell lines exposed to cabozantinib [42]. Thus, it is recommended that substrates of P-gp be used with caution when co-administered with cabozantinib to avoid a possible DDI associated with inhibition of this clinically significant transporter [10, 14, 15].

6 Conclusions Receptor tyrosine kinase inhibitor cabozantinib is approved for the treatment of progressive metastatic MTC (CometriqÒ) at a 140-mg FBE dose and for the treatment of

S. A. Lacy et al.

RCC following anti-angiogenic therapy (CabometyxTM) at a 60-mg FBE dose. The clinical pharmacokinetics of cabozantinib in HVs and cancer patients is characterized by a long plasma half-life, accumulation with daily dosing, and moderate variability. Although extensively metabolized, cabozantinib appears to be the principal, pharmacologically active circulating analyte. Results from clinical pharmacology studies have identified clinically relevant extrinsic and intrinsic factors that may affect cabozantinib plasma pharmacokinetics, including CYP3A4 inducers and inhibitors, a high-fat meal, and hepatic impairment. Exposure-response analyses suggest clinical anti-tumor activity is maintained in MTC patients following dose reductions from 140 to 60 mg FBE. Acknowledgments The authors wish to respectfully thank the medical professionals, patients and healthy volunteers that particpated in the clinical trials of cabozantinib reported in this manuscript. Compliance with Ethical Standards Funding The studies described in this manuscript were supported by Exelixis, Inc. Conflict of interest Steven A. Lacy is a stockholder and current employee of Exelixis, Inc. Dale R. Miles and Linh T. Nguyen were employees of Exelixis, Inc. when this work was performed. The authors contributed significantly to the design, conduct, analyses, and interpretation of the data, and were involved in the preparation, review, and approval of this manuscript. Ethical approval The clinical studies were conducted in accordance with the World Medical Association Declaration of Helsinki, the International Conference on Harmonisation Tripartite Guideline for Good Clinical Practice, and all applicable local regulations. Study protocols and informed consent documents were reviewed and approved by the institutional review board of participating institutions, and informed consent was obtained from all participants before any study-specified procedures were undertaken. All animal experiments were performed in facilities accredited by the Association for Assessment and Accreditation of Laboratory Care according to protocols approved by Institutional Animal Care and Use Committees.

References 1. Yakes FM, Chen J, Tan J, et al. Cabozantinib (XL184), a novel MET and VEGFR2 inhibitor, simultaneously suppresses metastasis, angiogenesis, and tumor growth. Mol Cancer Ther. 2011;10(12):2298–308. 2. Carmeliet P, Jain RK. Principles and mechanisms of vessel normalization for cancer and other angiogenic diseases. Nat Rev Drug Discov. 2011;10(6):417–27. 3. Trusolino L, Bertotti A, Comoglio PM. MET signaling: principles and functions in development, organ regeneration and cancer. Nat Rev Mol Cell Biol. 2010;11(12):834–48. 4. Birchmeier C, Birchmeier W, Gherardi E, et al. Met, metastasis, motility and more. Nat Rev Mol Cell Biol. 2003;4(12):915–25. 5. Comoglio PM, Giordano S, Trusolino L. Drug development of MET inhibitors: targeting oncogene addiction and expedience. Nat Rev Drug Discov. 2008;7(6):504–16.

6. Nickerson ML, Jaeger E, Shi Y, et al. Improved identification of von Hippel-Lindau gene alterations in clear cell renal tumors. Clin Cancer Res. 2008;14(15):4726–34. 7. Rini BI. Metastatic renal cell carcinoma: many treatment options, one patient. J Clin Oncol. 2009;27(19):3225–34. 8. Giubellino A, Linehan WM, Bottaro DP. Targeting the Met signaling pathway in renal cancer. Expert Rev Anticancer Ther. 2009;9(6):785–93. 9. Elisei R, Schlumberger MJ, Mu¨ller SP, et al. Cabozantinib in progressive medullary cancer. J Clin Oncol. 2013;31(29):3639–46. 10. CometriqTM (cabozantinib) capsules. US prescribing information. Exelixis, Inc. November 2012. 11. Choueiri TK, Pal SK, McDermott DF, et al. A phase 1 study of cabozantinib (XL184) in patients with renal cell carcinoma. Ann Oncol. 2014;25(8):1603–8. 12. Choueiri TK, Escudier B, Powles T, et al. Cabozantinib versus everolimus in advanced renal-cell carcinoma. N Engl J Med. 2015;373(19):1814–23. 13. Choueiri TK, Escudier B, Powles T, et al. Cabozantinib versus everolimus in advanced renal cell carcinoma (METEOR): final results from a randomized, open-label phase 3 trial. Lancet Oncol. 2016;17(7):917–27. 14. CabometyxTM (cabozantinib) tablets. US prescribing information. Exelixis, Inc. April 2016. 15. CabometyxTM (cabozantinib) tablets. Summary of product characteristics. Ipsen Pharma. September 2016. 16. ClinicalTrials.gov. Study of cabozantinib (XL184) vs placebo in subjects with hepatocellular carcinoma who have received prior sorafenib. ClinicalTrials.gov Identifier NCT01908426. Available from: https://clinicaltrials.gov/ct2/show/NCT01908426?term= cabozantinib&rank=10. Accessed 26 Sep 2016. 17. Kurzrock R, Sherman SI, Ball DW, et al. Activity of XL184 (cabozantinib), an oral tyrosine kinase inhibitor, in patients with medullary thyroid cancer. J Clin Oncol. 2011;29:2660–6. 18. Nguyen L, Holland J, Miles D, et al. Pharmacokinetic (PK) drug interaction studies of cabozantinib: effect of CYP3A4 inducer rifampin and inhibitor ketoconazole on cabozantinib plasma PK, and effect of cabozantinib on CYP2C8 probe substrate rosiglitazone plasma PK. J Clin Pharmacol. 2015;55(9):1012–23. 19. Nguyen L, Benrimoh N, Xie Y, Lacy S. Pharmacokinetics of cabozantinib tablet and capsule formulations in healthy adult subjects. Anticancer Drugs. 2016;27(7):669–78. 20. Nguyen L, Holland J, Ramies D, et al. Effect of renal and hepatic impairment on the pharmacokinetics of cabozantinib. J Clin Pharmacol. 2016;56(9):1130–40. 21. Nguyen L, Holland J, Mamelock R, et al. Evaluation of the effect of food and gastric pH on the single-dose plasma pharmacokinetics of cabozantinib in healthy adult subjects. J Clin Pharmacol. 2015;55:1293–302. 22. Lacy S, Hsu B, Miles D, et al. Metabolism and disposition of cabozantinib in healthy male volunteers and pharmacologic characterization of its major metabolites. Drug Metab Dispos. 2015;43:1190–207. 23. Miles D, Jumbe S, Lacy S, Nguyen L. Population pharmacokinetic model of cabozantinib in patients with medullary thyroid carcinoma and its application to an exposure-response analysis. Clin Pharmacokinet. 2016;55(1):93–105. 24. Miles DR, Wada DR, Jumbe NL, et al. Population pharmacokinetic/pharmacodynamic modeling of tumor growth kinetics in medullary thyroid cancer patients receiving cabozantinib. Anticancer Drugs. 2016;27(4):328–41. 25. Center for Drug Evaluation and Research. Clinical pharmacology and biopharmaceutics review[s] for cabozantinib (Cometriq), 2012a. Available from: www.accessdata.fda.gov/drugsatfda_ docs/nda/2012/203756Orig1s000ClinPharmR.pdf. Accessed 26 Sept 2016.

Clinical Pharmacokinetics and Pharmacodynamics of Cabozantinib 26. Bentzien F, Zuzow M, Heald N, et al. In vitro and in vivo activity of cabozantinib (XL184), an inhibitor of RET, MET, and VEGFR2, in a model of medullary thyroid cancer. Thyroid. 2013;23(12):1569–77. 27. Zhou L, Liu X-D, Sun M, et al. Targeting MET and AXL overcomes resistance to sunitinib therapy in renal cell carcinoma. Oncogene. 2016;35(21):2687–97. 28. Xiang Q, Chen W, Ren M, et al. Cabozantinib suppresses tumor growth and metastasis in hepatocellular carcinoma by dual blockade of VEGFR2 and MET. Clin Cancer Res. 2014;20(11):2959–70. 29. Sameni M, Tovar EA, Essenburg CJ, et al. Cabozantinib (XL184) inhibits growth and invasion of preclinical TNBC models. Clin Cancer Res. 2016;22(4):923–34. 30. Zhang L, Scorsone K, Woodfield SE, Page PE. Sensitivity of neuroblastoma to the novel kinase inhibitor cabozantinib is mediated by ERK inhibition. Cancer Chemother Pharmacol. 2015;76(5):977–87. 31. Song EK, Tai WM, Messersmith WA, et al. Potent antitumor activity of cabozantinib, a c-MET and VEGFR2 inhibitor, in colorectal cancer patient-derived tumor explant model. Int J Cancer. 2015;136(8):1967–75. 32. Cohen NA, Zeng S, Seifert AM, et al. Pharmacological inhibition of KIT activates MET signaling in gastrointestinal stromal tumors. Cancer Res. 2015;75(10):2061–70. 33. Sennino B, Ishiguro-Oonuma T, Wei Y, et al. Suppression of tumor invasion and metastasis by concurrent inhibition of c-Met and VEGF signaling in pancreatic neuroendocrine tumors. Cancer Discov. 2012;2(3):270–87. 34. Dai J, Zhang H, Karatsinides A, et al. Cabozantinib inhibits prostate cancer growth and prevents tumor-induced bone lesions. Clin Cancer Res. 2014;20(3):617–30.

35. European Medicines Agency. European Public Assessment Report (EPAR) for Cometriq, 2013. Available from: http://www. ema.europa.eu/docs/en_GB/document_library/EPAR_-_Public_ assessment_report/human/002640/WC500163705.pdf. Accessed 26 Sep 2016. 36. European Medicines Agency. Guideline on the investigation of bioequivalence. CPMP/EWP/QWP/1401/98 Rev. 1/Corr. London: European Medicines Agency; January 2010. 37. US Food and Drug Administration. Guidance for Industry. Bioavailability and bioequivalence studies for orally administered drug products: general considerations; revision 1. Rockville, MD: US Department of Health and Human Services, Food and Drug Administration; 2003. 38. US Food and Drug Administration. Guidance for industry. Bioavailability and bioequivalence studies submitted in NDAs and INDs: general considerations. Rockville, MD: US Department of Health and Human Services, Food and Drug Administration; 2014. 39. Foxx-Lupo WT, Sing S, Alwan L, Tykodi SS. A drug interaction between cabozantinib and warfarin in a patient with renal cell carcinoma. Clin Genitourin Cancer. 2016;14(1):119–21. 40. European Medicines Agency. Guideline on the investigation of drug interactions. London: European Medicines Agency; 2012. 41. US Food and Drug Administration. Guidance for industry. Drug interactions studies: study design, data analysis, implications for dosing, and labeling recommendations. Rockville, MD: US Food and Drug Administration; 2012. 42. Xiang QF, Zhang DM, Wang JN, et al. Cabozantinib reverses multidrug resistance of human hepatoma HepG2/adr cells by modulating the function of P-glycoprotein. Liver Int. 2015;35(3):110–23.