A first-in-human study of AMG 208, an oral MET inhibitor, in adult patients with advanced solid tumors.

BACKGROUND
This first-in-human study evaluated AMG 208, a small-molecule MET inhibitor, in patients with advanced solid tumors.


METHODS
Three to nine patients were enrolled into one of seven AMG 208 dose cohorts (25, 50, 100, 150, 200, 300, and 400 mg). Patients received AMG 208 orally on days 1 and days 4-28 once daily. The primary objectives were to evaluate the safety, tolerability, pharmacokinetics, and maximum tolerated dose (MTD) of AMG 208.


RESULTS
Fifty-four patients were enrolled. Six dose-limiting toxicities were observed: grade 3 increased aspartate aminotransferase (200 mg), grade 3 thrombocytopenia (200 mg), grade 4 acute myocardial infarction (300 mg), grade 3 prolonged QT (300 mg), and two cases of grade 3 hypertension (400 mg). The MTD was not reached. The most frequent grade ≥3 treatment-related adverse event was anemia (n = 3) followed by hypertension, prolonged QT, and thrombocytopenia (two patients each). AMG 208 exposure increased linearly with dose; mean plasma half-life estimates were 21.4-68.7 hours. One complete response (prostate cancer) and three partial responses (two in prostate cancer, one in kidney cancer) were observed.


CONCLUSIONS
In this study, AMG 208 had manageable toxicities and showed evidence of antitumor activity, particularly in prostate cancer.

AMG 208 is a small-molecule MET inhibitor with a 50% inhibitory concentration (IC 50 ) against wildtype MET of 5.2 nM. At higher concentrations, AMG 208 inhibited other kinases, such as VEGF receptor www.impactjournals.com/oncotarget 2 (VEGF-R2, IC 50 = 112 nM; data on file). AMG 208 suppressed proliferation and induced apoptosis in human tumor xenograft models (data on file).
We conducted a first-in-human study of AMG 208 to investigate its safety, tolerability, pharmacokinetics, and pharmacodynamics in patients with advanced solid tumors (ClinicalTrials.gov identifier: NCT00813384). We also evaluated antitumor activity and MET expression, amplification, and mutation status as potential biomarkers of response.

Safety and tolerability
Six patients had DLTs: 200 mg (n = 2), 300 mg (n = 2), and 400 mg (n = 2). In the 200-mg cohort, seven patients were initially enrolled, of whom two had a DLT (grade 3 increased aspartate aminotransferase [AST] and grade 3 thrombocytopenia), four completed without a DLT, and one withdrew early from the study due to disease progression. The protocol was amended to deescalate to 150 mg and then re-escalate to 200 mg ( Figure  1); three additional patients were then enrolled to the 200mg cohort, of whom two completed without a DLT, and Figure 1: Schematic overview of the dose escalation cohorts. Three to nine patients were enrolled into one of the following seven AMG 208 dose cohorts: 25, 50, 100, 150, 200, 300, and 400 mg. A standard 3+3 design was followed in cohorts 1-3, and a modified 3+3+3 design was followed in cohorts 4-7. The protocol was amended to evaluate an intermediate dose level of 150 mg after two DLTs (out of six patients) were observed with 200 mg AMG 208 (red arrow); re-escalation to 200 mg occurred after the 150-mg dose cohort was considered well tolerated (red arrow).
Oncotarget 18695 www.impactjournals.com/oncotarget one withdrew from the study due to stroke, which was not considered related to AMG 208. The 200-mg cohort was expanded to four additional patients, of whom three completed without a DLT, and one withdrew early from the study. Per protocol, two additional patients were allowed to enroll to the 200-mg cohort because of MET-positive status. In the 300-mg cohort, seven patients were initially enrolled, of whom two had a DLT (grade 4 acute myocardial infarction and grade 3 prolonged QT), four completed without DLT, and one withdrew early from the study. Three additional patients were enrolled to the 300mg cohort, and these three completed without a DLT. In the 400-mg cohort, eight patients were initially enrolled, of whom two had a DLT (both grade 3 hypertension), four completed without a DLT, and two withdrew early from the study. Three additional patients were enrolled to the 400-mg cohort, and these three completed without a DLT. A maximum tolerated dose (MTD) was not determined.
Seven patients reported treatment-related AEs considered serious: 200 mg (n = 3; increased AST, thrombocytopenia, and increased serum creatinine), 300 mg (n = 1; acute myocardial infarction), and 400 mg (n = 3; hypertension, pulmonary embolism, and neutropenia). Of these seven patients, three had treatment-related serious AEs leading to AMG 208 discontinuation: 200 mg (n = 1; increased AST), 300 mg (n = 1; acute myocardial infarction), and 400 mg (n = 1; hypertension). The increased AST was observed in a 43-year-old female with stage IV papillary renal cell carcinoma who developed grade 3 elevation of AST with grade 1 elevation of alanine aminotransferase (ALT) on day 8 of the study. AMG 208 was discontinued, and the AST and ALT were normalized within 1 week. The investigator did not report any relevant concomitant medications, and there were   Cardiac catheterization revealed ulcerated plaques in the right coronary artery and moderate/diffuse disease of the left anterior descending and circumflex arteries. The hypertension was observed in a 76-year-old male with stage IV prostate cancer who had a history of controlled hypertension. His blood pressure was elevated on days 7 to 10 of the study and ranged from 178/88 to 210/90. There were no other predisposing risk factors for elevated blood pressure except for the presence of bone pain before and during the event. AMG 208 was discontinued, and the event resolved within 4 days of discontinuation. Five patients died during the study, and none were considered treatment related (disease progression, n = 4; pulmonary hemorrhage, n = 1). The patient with pulmonary hemorrhage was diagnosed with pulmonary hypertension and pulmonary aspergillosis and had clear evidence of metastatic disease in the lung; thus, the grade 5 pulmonary hemorrhage was considered related to the disease.

Pharmacokinetics
AMG 208 pharmacokinetics was estimated for 53 patients who received ≥1 dose of AMG 208. After oral administration, linear increases of approximately 10-to 12-fold were observed over the 25-to 400-mg dose range in maximum concentration (C max ) and area under the concentration-time curve from time 0 to 24 hours (AUC 0-24h ) exposures. In the 400-mg cohort, mean C max and AUC 0-24h exposures on day 28 were 18.4 µg/mL and 245 µg•h/ mL, respectively. Mean plasma concentration-time profiles from each cohort are shown for 36 patients who completed 28 days of AMG 208 dosing ( Figure 2). AMG 208 mean plasma half-life estimates ranged from 21.4 to 68.7 hours and were consistent with a 1.81-to 3.43-fold accumulation observed after 28 days of repeated dosing.

Pharmacodynamics
The pharmacodynamic effects of AMG 208 on the following circulating biomarkers were evaluated: soluble MET, HGF, placental growth factor (PlGF), VEGF-R2, c-Kit, sFlt-1, VEGF, serum C-terminal telopeptide of type 1 collagen (sCTx), type 1 procollagen N-terminal propeptide (P1NP), and bone alkaline phosphatase (BALP). PlGF demonstrated a pharmacodynamic response to AMG 208; mean PlGF levels increased the most with the 400-mg dose at all time points (Figure 3). No other circulating biomarkers demonstrated a pharmacodynamic effect.  Figure  4A shows the change in the sum of the longest diameter for the best postdose response. There were one complete response (CR) and three partial responses (PRs). The CR was observed in a 66-year-old patient with prostate cancer at week 18, based on nontarget lesions evaluated by bone scans (300-mg cohort, Figure 5A); this patient was on the study for approximately 57 weeks. One confirmed PR was observed in a patient with kidney cancer at week 9 (200-mg cohort); this patient had a 33% tumor reduction and was on the study for approximately 23 weeks. Two unconfirmed PRs were observed in two patients with prostate cancer (both 400-mg cohort). One patient had a 33% tumor reduction and was on the study for   Oncotarget 18700 www.impactjournals.com/oncotarget scans. Eleven patients went off study early and had no follow-up scans, and one patient had incomplete coverage of target lesion at the follow-up scan and could not be evaluated. Thirty-three patients had stable disease, and nine patients had progressive disease.

Antitumor activity
Antitumor activity was further evaluated by 3'-deoxy-3'-18 F-fluorothymidine ( 18 F-FLT) positron emission tomography (PET) scans ( Figure 4B). Twenty-seven patients had evaluable baseline and week 5 18 F-FLT-PET scans. Seven patients had ≥25% reduction from baseline in 18 F-FLT uptake (maximum standardized uptake value [SUV max ]) at week 5 day 29, indicating a treatment response in tumor cell proliferation. Two of the seven proliferative responders (1 patient with prostate cancer [ Figure 5B] and 1 patient with kidney cancer) also had PR by RECIST 1.0 (33% reduction based on site read). Of the 10 patients with prostate cancer, three patients had a response per modified RECIST 1.0. As mentioned earlier, one patient with prostate cancer had a CR on bone scan (300-mg cohort). Two patients with prostate cancer had unconfirmed PRs based on CT scans, both in the 400-mg cohort. One patient had a 33% tumor reduction at week 17. The second patient had a 41% tumor reduction at week 27. Furthermore, three patients with prostate cancer were proliferative responders (36%, 41%, and 42% reductions in 18 F-FLT uptake).

Biomarkers
Twenty-two, 29, and 25 patients were analyzed for MET by immunohistochemistry (IHC), fluorescence in situ hybridization (FISH), and sequencing, respectively. No associations between response and MET testing were observed (Table 3).

DISCUSSION
In this first-in-human study of AMG 208, an oral MET inhibitor, the MTD was not reached at the highest dose tested (400 mg), and AMG 208 had manageable toxicities. The most common any grade AEs included fatigue and nausea; the most common grade ≥3 AE was anemia. DLTs were reported in six patients at the 200-to 400-mg dose levels.
Similar toxicities were observed with AMG 208 as seen with other multikinase VEGF/MET inhibitors. In clinical studies of cabozantinib, a multikinase MET and VEGF inhibitor, common AEs included fatigue, decreased appetite, and diarrhea [15,16], AEs also observed with AMG 208. Moreover, in a phase 2 study, foretinib, a dual MET/VEGF-R2 inhibitor, had a similar toxicity profile to AMG 208; common AEs included fatigue, hypertension, and gastrointestinal toxicities [17].
AMG 208 was orally bioavailable at the doses tested and exhibited a favorable pharmacokinetic profile. AMG 208 exposures increased linearly up to the 400-mg dose level, and mean estimates of elimination half-life ranged from 21.4 to 68.7 hours. After 28 days of once daily AMG 208 dosing, mean unbound trough concentrations ranged from 12.3 to 199 nM, thus exceeding the in vitro IC 50 estimate against wild-type MET (5.2 nM) at all dose levels. Among patients who received the 300-and 400mg AMG 208 doses, unbound trough concentrations approximated or slightly exceeded the in vitro IC 50 against VEGF-R2 (112 nM), which might explain higher increases in mean PIGF levels that were observed at the 400 mg dose level.
AMG 208 showed encouraging antitumor activity in prostate cancer, as suggested by the CR, 2 PRs, and 3 proliferative responses. In a phase 2 study of cabozantinib in prostate cancer, PFS was longer (23.9 versus 5.9 weeks with cabozantinib versus placebo), and 72% of evaluable patients had regression in soft tissue lesions with cabozantinib treatment [15]. It was postulated that cabozantinib's efficacy may be due to the simultaneous inhibition of MET and VEGF, and that the sole targeting of either the MET or VEGF pathways may not be sufficient in this disease. However, recent findings from the phase 3 study in castration-resistant prostate cancer showed no statistically significant improvement in OS with cabozantinib versus prednisone (median OS: 11 versus 9.8 months) [18], indicating that the hypothesis that efficacy in prostate cancer is due to dual inhibition of MET and VEGF may not be justified. AMG 208 has a different target coverage profile than cabozantinib; hence, lack of efficacy with cabozantinib in the prostate cancer setting may not be of relevance to AMG 208. . Furthermore, in another study, androgen deprivation was associated with a switch to MET signaling in prostate cancer cells [19]. All 10 patients with prostate cancer in the current study had previous androgen deprivation therapy.
Differences were observed between investigatorassessed and centrally assessed tumor responses; these may be attributable to inadequacies of RECIST 1.0, which considers bone and cystic lesions as nonmeasurable lesions. Some lesions defined as responders in the site reads per RECIST 1.0 were possibly bone or cystic lesions that the investigators considered as measurable lesions. Moreover, different lesions may have been measured in the site and central reads.
Patients selected by MET protein overexpression and MET amplification in gastroesophageal cancers [20,21], MET germline mutations in papillary renal cell cancer [17], and chromosome polyploidy in gastric cancer [22] have been associated with response to MET inhibitors. However, in our exploratory and retrospective analysis of MET in the study, no apparent associations between MET expression, amplification, and mutation status were observed. The results suggest that increased levels of MET expression by IHC did not correlate with response to MET inhibition as observed with AMG 208. One consideration is that some patients were treated during dose-escalation of AMG 208 and may not have received adequate doses to inhibit the pathway. Moreover, there may be a minimum threshold at which the MET/CEP7 ratio confers MET dependency and sensitivity to AMG 208, similar to recent reports of HER-2 and trastuzumab in gastric cancer [23]. Finally, these analyses were limited by the small sample size.
AMG 208 at 400 mg was the highest administered dose in this study. Although the MTD was not reached, the 400 mg dose was considered the recommended phase 2 dose. AMG 208 was initially investigated as a MET inhibitor. As mentioned earlier, the responses observed in prostate cancer are likely a result of its multikinase activity, similar to cabozantinib. The study was stopped before enrollment into the dose expansion phase. www.impactjournals.com/oncotarget It is hypothesized that inhibiting the MET pathway may overcome resistance to various therapies, including anti-EGFR inhibitors, platinum chemotherapy, and radiotherapy [24]. In a phase 1 study, tivantinib, a MET inhibitor and microtubule polymerization inhibitor [25], combined with erlotinib, an EGFR inhibitor, showed promising clinical activity with PR or stable disease observed in 15 of 32 patients [26]; in this same combination study, 6 of 8 patients with NSCLC achieved stable disease [26]. AMG 208 combined with other therapies has yet to be evaluated in clinical studies. Crosstalk between MET and other receptors, such as EGFR, HER2, integrin, and RON, may play a role in the development of resistance to targeted therapies, thus providing rationale to investigate MET pathway inhibitors in combination therapies.
MET continues to be an important target for cancer therapy. In addition to AMG 208, other small molecule MET inhibitors are under clinical development, some being selective inhibitors (eg, volitinib) and others being multikinase inhibitors (eg, crizotinib and cabozantinib) [24]. Several selective small molecule MET inhibitors have recently shown activity in early clinical trials, which include ABT-700 [27], AMG 337 [28], INC280 [29], MSC2156119J [30], SAR125844 [31], and volitinib [32]. Monoclonal antibodies targeting the MET pathway (eg, onartuzumab and rilotumumab) have also been evaluated in clinical trials [24]. Furthermore nanobodies to MET and indozoles have been tested in the preclinical setting as potential inhibitors of MET [33,34].
In conclusion, AMG 208 was well tolerated as monotherapy at doses up to 400 mg in patients with advanced solid tumors. Future studies evaluating MET pathway inhibitors, particularly in prostate cancer and/or in combination therapies, are warranted.

Eligibility criteria
Eligible patients (≥18 years) had a pathologically documented advanced solid tumor refractory to standard treatment or for which no standard therapy was available. Patients had measurable disease by RECIST version 1.0 [35]. Some patients with prostate cancer and nonmeasurable but evaluable disease (nontarget lesions only) were eligible (modified RECIST 1.0). Patients had an Eastern Cooperative Oncology Group (ECOG) performance status ≤2, life expectancy > 3 months, absolute neutrophil count ≥1.5 x 10 9 /L, platelets ≥100 x 10 9 /L, hemoglobin > 9 g/dL, serum creatinine < 2 mg/ dL, calculated creatinine clearance > 60 mL/min, ALT or AST < 3 x the upper limit of normal (ULN; if liver involvement was present, < 5 x ULN), total bilirubin 1.5 x ULN (if liver involvement was present, < 2 x ULN), alkaline phosphatase < 2 x ULN (if liver involvement or bone metastasis was present, ≤5 x ULN), and prothrombin or partial thromboplastin time < 1.5 x institutional ULN. Patients with primary central nervous system tumors or metastases were excluded. Each patient provided informed consent. Institutional review board approval was obtained for all study procedures.

Study design
This first-in-human, open-label study was to be conducted in two parts: (1) dose escalation and (2) dose expansion. In the dose escalation phase, 3-9 patients were enrolled into 1 of 7 dose cohorts (25,50,100,150,200, 300, and 400 mg) of AMG 208 (Figure 1). Patients were administered AMG 208 as a single oral dose followed by a 72-hour treatment-free period to evaluate pharmacokinetics. Beginning on day 4, patients received daily oral doses of AMG 208 up to day 28. If no DLT was seen on days 1-28, patients with no evident disease progression received AMG 208.
A DLT was defined as any grade ≥3 nonhematologic or grade 4 hematologic AE occurring during the first 28 days of treatment and possibly AMG 208 related. Treatment-related grade 3 thrombocytopenia could be considered a DLT if accompanied by grade ≥2 hemorrhage. Fatigue (unless grade 3 and lasting > 7 days or grade 4) and lymphopenia were not considered DLTs. For patients with liver involvement or bone metastases, DLTs did not include elevations in alkaline phosphatase unless > 8 x ULN when the baseline level was 2-5 x ULN. For patients with liver involvement, DLTs did not include elevations in AST or ALT unless > 8 x ULN and the baseline level was 3-5 x ULN. Serum creatinine > 2.5 mg/dL was considered a DLT.
In cohorts 1-3, a standard 3+3 design was followed. Enrollment into the next dose level occurred if no patients in the initial cohort experienced a DLT in the first 28 days of treatment. If a DLT occurred, the cohort was expanded to six patients. Enrollment into the next dose level occurred if no DLTs were observed. If ≥2 DLTs were observed among the six patients, enrollment was to be stopped. Following a report of two DLTs (out of six patients) at the 200-mg dose level, the protocol was amended to evaluate an intermediate dose level of 150 mg, and if well tolerated, re-escalation to the 200-mg dose level would occur. In cohorts 4-7, a modified 3+3+3 design was followed in order toprovide additional data to ascertain dose selection; if a second DLT was observed, three additional patients were to be enrolled at the same dose level for a total of at least nine patients. If no DLTs were observed, dose escalation to the next dose level would occur.
To enrich this study, patients with tumors with MET amplification or mutation (MET positive) were allowed www.impactjournals.com/oncotarget to enroll into the study at any time at the current dose escalation cohort if a slot was available. If the current escalation cohort was full, eligible MET-positive patients were assigned to the highest dose level deemed safe and well tolerated at the time of enrollment.

Study objectives
The primary objectives were to evaluate the safety, tolerability, and pharmacokinetics of AMG 208 and determine the MTD. Secondary objectives included evaluating tumor volume changes by CT or magnetic resonance imaging (MRI), decreases in tumor cell proliferation with 18 F-FLT-PET scanning, potential biomarkers that reflect MET target coverage, and associations among response and MET expression, amplification, and mutation.

Safety
Safety was evaluated based on AEs, vital signs, clinical laboratory measurements, electrocardiograms, and physical examinations. AEs were graded according to the Common Terminology Criteria for Adverse Events, version 3.0.

Pharmacokinetics
AMG 208 pharmacokinetics was evaluated after a single dose and after 28 days of repeated daily dose administrations. AMG 208 plasma concentrations were determined by liquid chromatography-tandem mass spectrometry from samples collected predose, 0.5, 1, 2, 4, 8, 24, and 48 or 72 hours after dosing on day 1, and pre-dose, 0.5, 1, 2, 4, 8, 24, and 48 hours on day 28. Pharmacokinetic parameters of observed C max , time of C max (t max ), and AUC 0-24h were determined by noncompartmental analysis using Phoenix WinNonlin software version 6.3 (Pharsight  , St. Louis, MO). AMG 208 accumulation was estimated as a ratio of AUC 0-24h on day 28 relative to day 1 for patients who remained on the study for ≥28 days.

Pharmacodynamics
Serum samples were collected on days 1 and 2; predose on days 8, 15, 22, 28; on day 64; and every 4 weeks thereafter and were analyzed for soluble MET, HGF, PlGF, VEGF-R2, c-Kit, sFlt-1, VEGF, sCTx, P1NP, and BALP. Pharmacodynamic effects were evaluated using time profiles with the mean ratio to baseline and standard error bars by treatment arms.
Levels of soluble MET, PlGF, VEGF, VEGF-R2, c-Kit, and sFlt-1 were quantified using multiplexed sandwich immunoassays with electrochemiluminescent detection (Meso Scale Discovery, Rockville, MD) following the manufacturer's instructions. Levels of HGF were analyzed using an analyte-specific enzymelinked immunosorbent assay (ELISA; R&D Systems, Minneapolis, MN) following the manufacturer's instructions and were compared to a standard curve. Samples were prepared and analyzed as previously described [36]. Levels of sCTx were quantified using the Serum Crosslaps ELISA (IDS Nordic, Herlev, Denmark) following the manufacturer's instructions. Levels of P1NP were quantified by radioimmunoassay (RIA) using the UniQ PINP RIA kit following the manufacturer's instructions and were compared to a standard curve (Covance Laboratories). Levels of BALP were quantified using the Access Ostase assay, a one-step immunoenzymatic assay, following the manufacturer's instructions (Beckman Coulter, Indianapolis, IN).

Antitumor activity
Tumor response was assessed by contrast-enhanced CT or MRI at screening; week 5; week 9; and every 8 weeks thereafter until disease progression and was evaluated by investigators based on modified RECIST version 1.0, which allowed some patients with metastatic prostate cancer and only nontarget bone lesions at screening to be evaluated with bone scans only. Response was also evaluated by an independent central imaging laboratory. Proliferative response was assessed by 18 F-FLT-PET scans at baseline and week 5. An antiproliferative response was defined as ≥25% reduction in 18 F-FLT SUV max .

Biomarkers
Archival formalin-fixed, paraffin-embedded (FFPE) tumor samples were analyzed for membrane and cytoplasmic MET expression by IHC at Mosaic Laboratories (Lake Forest, CA). Tumors were sectioned and stained with hematoxylin and eosin to identify regions of high tumor-cell content. IHC was performed using an anti-MET antibody (goat IgG, polyclonal, clone AF276) from R&D Systems. Staining was evaluated by a trained pathologist, and expression was evaluated for cellular localization of staining, intensity, subcellular localization, and percentage of tumor cells staining positive. Staining was evaluated on a semi-quantitative scale, and the percentage of tumor cells staining at the following intensities was recorded: 0 (unstained), 1+ (weak staining), 2+ (moderate staining), and 3+ (strong staining). An H-score was calculated based on the percentage of cells stained at each intensity as follows: (3 x percent cells staining at 3+) + (2 x percent cells staining at 2+) + (1 x percent cells staining at 1+).
Archival FFPE tumor samples were analyzed for www.impactjournals.com/oncotarget MET amplification by FISH at Histogenex Laboratories (Antwerp, Belgium). Tumors were pretreated and stained with MET/SE7 Probe kit (KBI-10719, Kreatech, Durham, NC) using a VP2000 autostainer and a Thermobrite Hybrid system (Abbott Molecular, Des Plaines, IL) following the manufacturers' instructions. Twenty nuclei were evaluated per sample, and the quantification of MET and SE7 signals was used to calculate the average MET copies per nuclei, average centromeric copies per nuclei, and the ratio of MET copies to SE7 copies. DNA was extracted from archival FFPE tumor samples and interrogated for MET mutations using the SURVEYOR Nuclease assay (Transgenomic, Inc., Omaha, NE) followed by Sanger sequencing.

Statistical analysis
Descriptive statistics on continuous data included means, medians, standard deviations, and ranges. Categorical data were summarized by frequency counts and percentages. All patients who received ≥1 dose of AMG 208 were included in the safety analyses. Treatmentemergent and treatment-related AEs occurring after the initial dose of AMG 208 and before the end of the study or 30 days after the last dose of AMG 208 (whichever occurred later) were presented descriptively. No formal analysis was done for efficacy. An exploratory summary of tumor response was produced. Descriptive analyses for biomarkers and associations between biomarkers and efficacy were produced. Results from the primary analysis (date of data cutoff: December 17, 2012) are presented.