This study was an open-label, single-arm, phase II clinical trial evaluating the efficacy of selumetinib in advanced or metastatic HCC. The study was performed by the Southeastern Phase II Consortium and the Ohio State University Phase II Consortium. The human participants committees at each participating center approved this study, and all patients provided written informed consent before participation. All trial procedures were conducted in accordance with the principles established by the Helsinki Declaration.
Patients enrolled on this study had either histologically proven or α-fetoprotein (AFP > 1,000 ng/dL) –confirmed HCC. They could not be considered candidates for potentially curative therapies. Prior regional therapy or ablative therapy was allowed, but prior systemic therapy (including sorafenib) was not. History of liver transplantation was exclusionary (CONSORT, ).
Eastern Cooperative Oncology Group (ECOG) performance status ≤ 1 was required in addition to the following laboratory criteria: leukocytes greater than 3,000/μL, absolute neutrophil count greater than 1,500/μL, platelets greater than 75,000/μL, total bilirubin less than two times the upper limit of normal (ULN), AST/ALT less than five times institutional ULN, creatinine less than 1.5 mg/dL (or creatinine clearance greater than 60 mL/min), and INR less than 1.4.
If cirrhosis was present, the patient had to meet criteria for Child-Pugh class A or B. If Child-Pugh B cirrhosis was present, the patient could not have significant encephalopathy or ascites that required ongoing paracentesis, and the patient had to meet the stated laboratory criteria.
Selumetinib was formulated as a mix and drink preparation by using captisol as the diluent. Dosing was 100 mg orally twice per day to patients who had fasted for a minimum of 2 hours before dosing. Days 1 through 3 of the study constituted the initial PK analysis; the patient took a single morning dose of 100 mg selumetinib followed by PK sampling. After the 48-hour PK sample, patients began twice-daily therapy continuously. A cycle of therapy was 21 days of therapy.
Dose modifications were planned for any grade 3 or 4 toxicities. Up to two dose reductions were allowed; the first was a reduction to 50 mg twice per day, and the second was a reduction to 50 mg once daily. Rash was managed with interruption of therapy for intolerable grade 2 or greater severity followed by dose reduction after resolution to grade 1 or tolerable grade 2.
PK sampling was performed for the first 15 patients on days 1 through 3 (single-dose PK) and days 15 to 17. Tumor biopsies were performed before day 1 with a 14-gauge to 20-gauge needle and up to four passes. Biopsies were immediately frozen in liquid nitrogen and were not used for diagnostic purposes. Paired tumor biopsies were obtained in the first five patients; subsequently, funding issues required us to amend the protocol to obtain only a baseline tumor biopsy. Analyses presented are based on the paired biopsy samples. Imaging for tumor response was performed after every two cycles.
Tumor Biopsy and Western Blot Analysis
Patients who consented to an optional biopsy had paired tumor biopsies performed before treatment and during treatment with selumetinib 100 mg orally twice per day on day 7, approximately 2 hours after selumetinib dose (presumed to be at steady-state). Specimens were snap frozen in liquid nitrogen and stored at −80°C until they were additional processed. Weighted tumor samples were mixed with the appropriate amount of the T-PER tissue protein extraction reagent (Thermo Scientific, catalog No. 78,510; Rockford, IL) containing protease inhibitor cocktail, 2 mmol/L phenylmethyl sulfonyl fluoride, 2 mmol/L Na3VO4, and 6.4 mg/mL p-nitrophenlyphosphate. Samples were homogenized with the polymerase chain reaction tissue homogenizing kit (Fisher Scientific, Rockford, IL). The homogenate was centrifuged at 13,000 × g for 30 minutes at 4°C, the supernatant was collected, and the protein concentration was determined by the Bradford assay. Proteins were separated by sodium dodecyl sulfate –polyacrylamide gel electrophoresis and were transferred to nitrocellulose membranes, which were then blotted with phospho-p44/42 map kinase (p-ERK; Thr202/Tyr204; 9101; Cell Signaling, Danvers, MA), p44/42 MAP kinase (9102; Cell Signaling), phospho-MEK1/2 (Ser 217/221; 9121; Cell Signaling), MEK1/2 (9122; Cell Signaling), phospho-p38 MAPK (Thr180/Tyr182; 12F8; 4631, Cell Signaling), p38 MAP kinase (9212, Cell Signaling), phospho-Akt (Ser473; 9271; Cell Signaling), Akt1/2 (N-19; SC-1619, Santa Cruz Biotechnology, Santa Cruz, CA), phospho-STAT3 (Tyr 705; 9131L; Cell Signaling), STAT3 (F-2; Sc-8019; Santa Cruz Biotechnology, Santa Cruz, CA), GAPDH (MMS-580S; Covance, Princeton, NJ).
Blood samples for determination of selumetinib and N-desmethyl selumetinib were taken on day 1 and day 15. The PK analyses were performed at Clinical Pharmacology and DMPK, Alderley Park, AstraZeneca, United Kingdom. The PK variables for the patients with comprehensive PK sampling were estimated by noncompartmental analysis by using WINNonlin (version 5.2, Scientific Consultant, Apex, NC). The actual sampling time, as opposed to the protocol-scheduled time, was used in the derivation of PK parameters.
The following parameter estimates were estimated for selumetinib from the observed concentration-time profiles: maximum plasma concentration (Cmax), the Cmax at steady-state (Css max), the time of maximum concentration (tmax), and the tmax at steady-state (tss max). Parameters were determined by inspection of the concentration-time profiles.
The apparent clearance (CL/F after the single dose and CLss/F after multiple dosing) were determined from the ratio of dose to area under the concentration curve (AUC) or dose to AUCss. Apparent volume of distribution (Vz/F) after oral dosing was calculated by dividing the dose by the product of λz × AUC. The estimated volume of distribution at steady-state after oral dosing (Vss/F) was determined from the mean residence time (MRT) × CL/F.
The accumulation ratio (RAC) was calculated as the ratio of the AUCss and AUC(0-12) on day 1. The AUC(0-12) was determined for N-desmethyl selumetinib on days 1 and 15 to enable the calculation of the metabolite-to-parent percentage on days 1 and 15 (N-desmethyl selumetinib AUC(0-12) ÷ selumetinib AUC(0-12) × 100). The time dependency (Tc or linearity factor) of the pharmacokinetics on multiple dosing was assessed by the calculation of the ratio of AUCss to AUC on day 1.
The primary end point was the objective response rate. Secondary end points included the time to event functions of progression, progression-free survival (PFS), and overall survival (OS). For sample size calculation, an optimal two-stage design18
was used. The information used in the calculations of this design were the values of P0
= .03, P1
= .15, α= .1, and power = 90%.
In the first stage, 19 patients were entered, with the assumption that at least 17 would be eligible. If there was at least one response among these 17 eligible patients, an additional 25 patients would have been entered for the second stage, of whom 22 would be assumed to be eligible. Thus, a total of 44 patients could have been enrolled on the study. The probability of early stopping was 0.60 if the true response rate was 0.03, and it was 0.06 if the true response rate was 0.15. Moreover, the overall probability of rejecting the treatment was 0.90 if the true response rate was 0.03, and it was 0.10 if the true response rate was 0.15.
The Kaplan-Meier (or product-limit) method was used to estimate all time-to-event functions. Time to disease progression (TTP) has been defined as the time from the start of treatment to disease progression. Deaths occurring in the absence of proven disease progression were censored. PFS has been defined as time from the start of treatment to disease progression or death as a result of any cause. OS has been defined as time from the start of treatment to death as a result of any cause. Exact 95% CIs were calculated for each proportion of interest. These proportions have been reported as percentages. Statistical analyses were performed with SAS statistical software (version 9.2, SAS Institute, Cary, NC). PK analysis was performed with software noted in the Methods section, and parameter estimates were summarized with descriptive statistics.