A Trial of Lenvatinib (E7080) in Subjects With Iodine-131 Refractory Differentiated Thyroid Cancer to Evaluate Whether an Oral Starting Dose of 18 Milligram (mg) Daily Will Provide Comparable Efficacy to a 24 mg Starting Dose, But Have a Better Safety Profile

Eisai Inc.241 enrolled

Overview

This is a multicenter, randomized, double-blind study being conducted as a postmarketing requirement to the US Food and Drug Administration (FDA) to evaluate whether there is a lower starting dosage of lenvatinib 24 mg once daily (QD) that provides comparable efficacy but has a better safety profile in participants with radioiodine-refractory differentiated thyroid cancer RR-DTC with radiographic evidence of disease progression within the prior 12 months.

This study consists of 2 phases, the Prerandomization Phase and the Randomization Phase. The Prerandomization Phase will last no longer than 28 days and will include a Screening Period to establish protocol eligibility and a Baseline Period to confirm eligibility and establish disease characteristics prior to randomization and treatment. The Randomization Phase will consist of a Treatment Period and a Follow-up Period. It will begin at the time of randomization of the first participant and will consist of 28-day blinded study treatment cycles. The data cutoff for the primary analysis will occur at the end of the Randomization Phase, which is defined as when the last participant enrolled completes the Week 24 tumor assessments or discontinues study treatment if before Week 24. Participants on treatment at the time of data cutoff for the primary analysis will remain on blinded investigational product until the primary analysis has been completed, after which they will transition to commercial lenvatinib outside the study. The last participant's last assessment (Off-treatment visit) will be the End of Study. Participants will be randomly assigned to treatment with 1 of 2 blinded starting dosages of lenvatinib in a 1:1 ratio to receive lenvatinib 18 mg or 24 mg orally once daily (QD). Dose reductions will occur in succession based on the previous dose level (24, 20, 14, 10, and 8 mg QD, or 18, 14, 10, 8, and 4 mg QD, respectively). Once the dose has been reduced, it may not be increased at a later date.

Study Type

INTERVENTIONAL

Allocation

RANDOMIZED

Purpose

TREATMENT

Masking

Eligibility

Sex: ALLMin age: 18 Years

Medical Language ↔ Plain English

Inclusion Criteria: 1. Participants must have histologically or cytologically confirmed diagnosis of one of the following differentiated thyroid cancer (DTC) subtypes: 1. Papillary thyroid cancer (PTC) * Follicular variant * Variants (including but not limited to tall cell, columnar cell, cribriform-morular, solid, oxyphil, Warthin's-like, trabecular, tumor with nodular fasciitis-like stroma, Hürthle cell variant of papillary carcinoma, poorly differentiated) 2. Follicular thyroid cancer (FTC) * Hurthle cell * Clear cell * Insular 2. Measurable disease meeting the following criteria and confirmed by central radiographic review: 1. At least 1 lesion of greater than or equal to (\>=)1.0 centimeter (cm) in the longest diameter for a non-lymph node or \>=1.5 cm in the short-axis diameter for a lymph node that is serially measurable according to Response Evaluation Criteria in Solid Tumors (RECIST 1.1) using computed tomography/magnetic resonance imaging (CT/MRI). If there is only 1 target lesion and it is a non-lymph node, it should have a longest diameter of \>=1.5 cm. 2. Lesions that have had external beam radiotherapy (EBRT) or locoregional therapies such as radiofrequency (RF) ablation must show evidence of progressive disease based on RECIST 1.1 to be deemed a target lesion. 3. Participants must show evidence of disease progression within 12 months (an additional month will be allowed to accommodate actual dates of performance of screening scans, that is, within less than or equal to \<=13 months) prior to signing informed consent, according to RECIST 1.1 assessed and confirmed by central radiographic review of CT and/or MRI scans. 4. Participants must be Iodine-131 refractory/resistant as defined by at least one of the following: 1. One or more measurable lesions that do not demonstrate iodine uptake on any radioiodine scan. 2. One or more measurable lesions that have progressed according to RECIST 1.1 within 12 months (an additional month will be allowed to accommodate actual dates of performance of screening scans, ie, within ≤13 months) after Iodine-131 therapy, despite demonstration of radioiodine avidity at the time of that treatment by pre- or posttreatment scanning. These participants must not be eligible for possible curative surgery. 3. Cumulative activity of Iodine-131 of greater than (\>) 600 millicurie (mCi) or 22 gigabecquerels (GBq), with the last dose administered at least 6 months prior to study entry. 5. Participants with known brain metastases who have completed whole brain radiotherapy, stereotactic radiosurgery or complete surgical resection, will be eligible if they have remained clinically stable, asymptomatic, and off steroids for one month. 6. Participants must be receiving thyroxine suppression therapy and thyroid stimulating hormone (TSH) should not be elevated (TSH should be \<=5.50 micro-international units per liter \[mcIU/ML\]). When tolerated by the participant, thyroxine dose should be changed to achieve TSH suppression (TSH \<0.50 mcIU/ML) and this dose may be changed concurrently upon starting study drug treatment. 7. All chemotherapy or radiation-related toxicities must have resolved to Grade \<2 severity per Common Terminology Criteria for Adverse Events (CTCAE v4.03), except alopecia and infertility. 8. Participants must have an Eastern Cooperative Oncology Group (ECOG) Performance Status of 0, 1, or 2. 9. Adequately controlled blood pressure (BP) with or without antihypertensive medications, defined as BP ≤150/90 mmHg at Screening and no change in antihypertensive medications within 1 week prior to Cycle 1/Day 1. 10. Adequate renal function defined as calculated creatinine clearance \>=30 mL/min per the Cockcroft and Gault formula. 11. Adequate bone marrow function: 1. Absolute neutrophil count (ANC) \>=1500/mm\^3 (\>=1.5\*10\^3/uL) 2. Platelets \>=100,000/mm\^3 (\>=100\*10\^9/L) 3. Hemoglobin \>=9.0 g/dL 12. Adequate blood coagulation function as evidenced by an International Normalized Ratio (INR) \<=1.5. 13. Adequate liver function: 1. Bilirubin \<=1.5\*upper limit of normal (ULN) except for unconjugated hyperbilirubinemia or Gilbert's syndrome. 2. Alkaline phosphatase, alanine aminotransferase (ALT), and aspartate aminotransferase (AST) \<=3\*ULN (\<=5\*ULN if participant has liver metastases). If alkaline phosphatase is \>3\*ULN (in absence of liver metastases) or \>5\*ULN (in presence of liver metastases) AND the participant also is known to have bone metastases, the liver-specific alkaline phosphatase must be separated from the total and used to assess the liver function instead of total alkaline phosphatase. 14. Males or females age \>=18 years at the time of informed consent. 15. Females must not be lactating or pregnant at Screening or Baseline (as documented by a negative beta-human chorionic gonadotropin \[B-hCG\] test with a minimum sensitivity of 25 IU/L or equivalent units of B-hCG. A separate baseline assessment is required if a negative screening pregnancy test was obtained more than 72 hours before the first dose of study drug. 16. All females will be considered to be of childbearing potential unless they are postmenopausal (amenorrheic for at least 12 consecutive months, in the appropriate age group and without other known or suspected cause) or have been sterilized surgically (ie, bilateral tubal ligation, total hysterectomy or bilateral oophorectomy, all with surgery at least 1 month before dosing). 17. Females of childbearing potential should avoid becoming pregnant and use highly effective contraception while on treatment with lenvatinib and for at least 1 month after finishing treatment. Females of childbearing potential must not have had unprotected sexual intercourse within 30 days before study entry and must agree to use a highly effective method of contraception (example, total abstinence, an intrauterine device, a contraceptive implant, an oral contraceptive, or have a vasectomized partner with confirmed azoospermia) throughout the entire study period and for 30 days after study drug discontinuation. Females who are using hormonal contraceptives must have been on a stable dose of the same hormonal contraceptive product for at least 4 weeks before dosing and must continue to use the same contraceptive during the study and for 30 days after study drug discontinuation. Women using oral hormonal contraceptives should add a barrier method. 18. Participants must voluntarily agree to provide written informed consent. 19. Participants must be willing and able to comply with all aspects of the protocol. Exclusion Criteria: 1. Anaplastic or medullary carcinoma of the thyroid. 2. Diagnosed with meningeal carcinomatosis. 3. Two or more prior vascular endothelial growth factor (VEGF)/vascular endothelial growth factor receptor (VEGFR)-targeted therapies or any ongoing treatment for RR-DTC other than TSH-suppressive thyroid hormone therapy. 4. Prior treatment with lenvatinib. 5. Participants who have received any anticancer treatment within 21 days or any investigational agent within 30 days (or 5 half-lives) prior to the first dose of study drug and should have recovered from any toxicity related to previous anticancer treatment. This does not apply to the use of TSH suppressive thyroid hormone therapy. 6. Major surgery (example, laparotomy, thoracotomy or joint replacement) within 3 weeks prior to randomization or elective surgery scheduled to be performed during the study. 7. Participants having \>1+ proteinuria on urine dipstick testing will undergo 24-hour urine collection for quantitative assessment of proteinuria. Participants with urine protein \>=1 g/24 h will be ineligible. 8. Gastrointestinal malabsorption or any other condition that in the opinion of the investigator might affect the absorption of study drug. 9. Significant cardiovascular impairment: history of congestive heart failure greater than New York Heart Association (NYHA) Class II, unstable angina, myocardial infarction or cerebral vascular accident within 6 months of the first dose of study drug, or cardiac arrhythmia associated with hemodynamic instability. 10. Prolongation of corrected QT interval (QTc) to \>480 ms as demonstrated by a repeated electrocardiogram (ECG) or a clinically significant ECG abnormality, including a marked prolonged QT/QTc interval (example, a repeated demonstration of a QTc interval \>500 ms). 11. Active hemoptysis (bright red blood of at least 0.5 teaspoon) within 3 weeks prior to the first dose of study drug. 12. Active infection (any infection requiring treatment). 13. Active malignancy (except for DTC or definitively treated melanoma in-situ, basal or squamous cell carcinoma of the skin, or carcinoma in-situ of the cervix) within the past 24 months. 14. Bleeding or thrombotic disorders. 15. Known intolerance to study drug (or any of the excipients). 16. Any medical or other condition that in the opinion of the investigator(s) would preclude the participant's participation in a clinical study. 17. Females who are pregnant or breastfeeding.

Outcomes

Primary Outcomes

Objective Response Rate (ORR) as of Week 24 (ORR24wk)

ORR as of Week 24 was defined as the percentage of participants with best overall response (BOR) of complete response (CR) or partial response (PR) as of the Week 24 time point or earlier, as measured by Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST 1.1). CR: was disappearance of all target lesions. Any pathological lymph nodes (target or non-target) had to be reduced in short axis to less than (\<) 10 millimeter (mm). PR: was at least a 30 percent (%) decrease in sum of diameter (SOD) of target lesions, taking as reference the baseline SOD.

Time frame: From the date of randomization up to Week 24

Percentage of Participants With Grade 3 or Higher Treatment-emergent Adverse Events (TEAEs) in the First 24 Weeks

This outcome measure reports TEAEs in the first 24 weeks only. A TEAE was defined as any adverse event (AE) that had an onset date on or after the first dose of study drug up to 28 days following the last dose of study drug, or a worsening in severity from Baseline (pretreatment). In addition, if an AE reemerged during treatment, having been present at pretreatment (Baseline) but stopped before treatment, it was also counted as a TEAE. A severity grade was defined by the Common Terminology Criteria for Adverse Events (CTCAE) Version 4.03. As per CTCAE, Grade 1 scales as Mild; Grade 2 scales as Moderate; Grade 3 scales as severe or medically significant but not immediately life threatening; Grade 4 scales as life-threatening consequences; and Grade 5 scales as death related to AE.

Time frame: Baseline up to Week 24

Secondary Outcomes

Progression-free Survival (PFS)

PFS, defined as the time from the date of randomization to the date of first documentation of PD, or date of death, whichever occurs first, as measured by RECIST V1.1. PD: 20% increase in the sum of the pertinent diameters (SOD) of target lesions, taking as reference the smallest sum SOD recorded since the treatment started or the appearance of one or more new lesions. PFS was analyzed using the Kaplan-Meier method. As planned, data for this endpoint was analyzed and collected till Primary completion date.

Time frame: Time from the date of randomization to the date of first documentation of PD, or date of death, whichever occurs first up to approximately 2 years 6 months

PFS After Next Line of Treatment (PFS2)

PFS2, defined as the time from randomization to PD on next-line treatment, or death from any cause, whichever occurred first, as measured by RECIST V1.1. PD: 20% increase in the SOD of target lesions, taking as reference the smallest sum SOD recorded since the treatment started or the appearance of one or more new lesions. PFS was analyzed using the Kaplan-Meier method. As planned, data for this endpoint was analyzed and collected till Primary completion date.

Time frame: Time from randomization to PD on next-line treatment or death from any cause, whichever occurs first up to approximately 2 years 6 months

Number of Participants With TEAE and Serious Adverse Events (SAEs)

TEAEs were defined as those AEs that occurred (or worsened, if present at Baseline) after the first dose of study drug through 28 days after the last dose of study drug. An AE was defined as any untoward medical occurrence in a participants or clinical investigation participant administered an investigational product. An AE does not necessarily have a causal relationship with medicinal product. SAE was defined as any AE if it resulted in death or life-threatening AE or required inpatient hospitalization or prolongation of existing hospitalization or resulted in persistent or significant incapacity or substantial disruption of the ability to conduct normal life functions or was a congenital anomaly/birth defect.

Time frame: From date of first administration of study drug up to 28 days after last dose of study drug up to approximately 3 years 3 months

Time to Treatment Discontinuation Due to an Adverse Event (AE)

Time to Treatment Discontinuation due to an AE (such as abdominal distention, appendicitis perforated, arthralgia, anemia, etc) was analyzed using the Kaplan-Meier method. As planned, data for this endpoint was analyzed and collected till Primary completion date.

Time frame: From date of first administration of study drug up to approximately 2 years 6 months

Number of Dose Reductions

Number of dose reduction was reported as number of participants who underwent one or more number of dose reductions. As planned, data for this endpoint was analyzed and collected till Primary completion date.

Time frame: From date of first administration of study drug up to approximately 2 years 6 months

Time to First Dose Reduction

Time to First Dose Reduction was analyzed using the Kaplan-Meier method. As planned, data for this endpoint was analyzed and collected till Primary completion date.

Time frame: From date of first administration of study drug up to approximately 2 years 6 months

Model Predicted Apparent Total Clearance (CL/F) Following Oral Dosing of Lenvatinib

Sparse pharmacokinetic (PK) samples (approximately 9 per participant) were collected and analyzed using a population PK approach to estimate PK parameters. Lenvatinib total plasma concentration data were pooled with data from studies E7080-G000-303 (NCT01321554) and E7080-G000-201 (NCT00784303), and a population PK model was applied to the pooled dataset. Individual predicted CL/F for lenvatinib was then derived from the PK model by starting dose.

Time frame: Cycle 1 Day 1: 0.5-4 hours and 6-10 hours postdose; Cycle 1 Day 8: predose, Cycle 1 Day 15: predose, 0.5-4 hours and 6-10 hours postdose; Cycle 1 Day 22: optionally at predose; Cycle 2 Day 1: predose and 2-12 hours postdose (Cycle length=28 days)

Model Predicted Area Under the Plasma Drug Concentration-time Curve (AUC) for Lenvatinib

Sparse PK samples (approximately 9 per participant) were collected and analyzed using a population PK approach to estimate PK parameters. Lenvatinib total plasma concentration data were pooled with data from studies E7080-G000-303 (NCT01321554) and E7080-G000-201 (NCT00784303), and a population PK model was applied to the pooled dataset. Individual predicted AUC for lenvatinib was then derived from the PK model by starting dose.

Parameter Estimates From the Population Pharmacokinetic/Pharmacodynamic (PK/PD) Model Describing the Relationship Between Lenvatinib Exposure (AUC) and Thyroglobulin Levels

The relationship between exposure to lenvatinib and change from baseline in thyroglobulin was planned to be analyzed using a model-based approach.

Parameter Estimates From the Population Pharmacokinetic/Pharmacodynamic (PK/PD) Model Describing the Relationship Between Lenvatinib Exposure (AUC) and Thyroid-Stimulating Hormone (TSH) Levels

The relationship between exposure to lenvatinib and change from baseline in TSH was planned to be analyzed using a model-based approach.

Baseline Level Estimates From the Population PK/PD Model Describing the Relationship Between Lenvatinib Exposure (AUC) and Vascular Endothelial Growth Factor (VEGF), Soluble Tie-2, Angiopoietin-2 (Ang-2) and Fibroblast Growth Factor-23 (FGF23) Levels

Per the planned population PK/PD analysis for this endpoint, Arms/Groups were combined and lenvatinib total plasma concentration and serum biomarker data for VEGF, Ang-2, soluble Tie-2, and FGF23 from this study were combined with data from study E7080-G000-303 (NCT01321554). The relationship between lenvatinib exposure at the time of measurement of biomarker was described using PK/PD modelling. Initially, PK/PD models were developed individually for each biomarker and then combined into a single combined model. Changes in biomarker levels over time related to lenvatinib exposure were best described by an indirect response, sigmoidal Emax model. For the final combined model, baseline level estimates were determined separately for each biomarker. The data presented are the model predicted baseline estimates, with Measure Type "Number." They are population PK/PD model predictions and have been estimated using non-linear mixed effects modelling.

Mean Residence Time (MRT) Estimates From the Population PK/PD Model Describing the Relationship Between Lenvatinib Exposure (AUC) and VEGF, Soluble Tie-2, Ang-2 and FGF23 Levels

Per the planned population PK/PD analysis for this endpoint, Arms/Groups were combined and lenvatinib total plasma concentration and serum biomarker data for VEGF, Ang-2, soluble Tie-2, and FGF23 from this study were combined with data from study E7080-G000-303 (NCT01321554). The relationship between lenvatinib exposure at the time of measurement of biomarker was described using PK/PD modelling. Initially, PK/PD models were developed individually for each biomarker and then combined into a single combined model. Changes in biomarker levels over time related to lenvatinib exposure were best described by an indirect response, sigmoidal Emax model. For the final combined model, MRT estimates were determined separately for each biomarker. The data presented are the model predicted MRT estimates, with Measure Type "Number." They are population PK/PD model predictions and have been estimated using non-linear mixed effects modelling.

Hill Coefficient Estimates From the Population PK/PD Model Describing the Relationship Between Lenvatinib Exposure (AUC) and VEGF, Soluble Tie-2, Ang-2 and FGF23 Levels

Per the planned population PK/PD analysis for this endpoint, Arms/Groups were combined and lenvatinib total plasma concentration and serum biomarker data for VEGF, Ang-2, soluble Tie-2, and FGF23 from this study were combined with data from study E7080-G000-303 (NCT01321554). The relationship between lenvatinib exposure at the time of measurement of biomarker was described using PK/PD modelling. Initially, PK/PD models were developed individually for each biomarker and then combined into a single combined model. Changes in biomarker levels over time related to lenvatinib exposure were best described by an indirect response, sigmoidal Emax model. For the final combined model, Hill Coefficient estimates were determined separately for each biomarker. The data presented are the model predicted Hill Coefficient estimates, with Measure Type "Number." They are population PK/PD model predictions and (\&) have been estimated using non-linear mixed effects modelling.

Parameter Estimates From the PK/PD Model for Tumor Growth Inhibition and Serum Biomarkers Tie-2 and Ang-2

Per the planned analysis, Arms/Groups were combined \& tumor-growth inhibition models based on lenvatinib \& placebo data from this study combined with study NCT01321554. Effects of tumor growth rate, drug effects, tumor resistance, \& tumor size reduction related to biomarker response were assessed. Longitudinal data of sum of the longest diameter for target lesion by investigator assessment in this study \& independent reviewer assessment in study NCT01321554 were used. Changes in Ang-2 \& soluble Tie-2 were evaluated, individually \& in combination for their impact on tumor size. The concomitant use of lenvatinib \& biomarker changes due to drug effects as predictors of tumor size were also evaluated. The final integrated model for tumor growth \& biomarkers included effects of lenvatinib exposure \& tumor growth reduction related to Tie-2 \& Ang-2 biomarkers as significant predictors. Data presented are the parameters defining this non-linear mixed effects model, with Measure Type "Number."

Time frame: Baseline up to week 120

Lenvatinib Mean AUC Resulting in 50% of the Emax (EC50) Estimate From the PK/PD Model for Tumor Growth Inhibition and Serum Biomarkers Tie-2 and Ang-2

Per the planned analysis, Arms/Groups were combined \& tumor-growth inhibition models based on lenvatinib \& placebo data from this study combined with study NCT01321554. Effects of tumor growth rate, drug effects, tumor resistance, \& tumor size reduction related to biomarker response were assessed. Longitudinal data of the sum of the longest diameter for target lesion by investigator assessment in this study \& independent reviewer assessment in study NCT01321554 were used. Changes in Ang-2 \& soluble Tie-2 were evaluated, individually \& in combination for their impact on tumor size. The concomitant use of lenvatinib \& biomarker changes due to drug effects as predictors of tumor size were also evaluated. The final integrated model for tumor growth \& biomarkers included effects of lenvatinib exposure \& tumor growth reduction related to Tie-2 \& Ang-2 biomarkers as significant predictors. Data presented are EC50 estimated using non-linear mixed effects modeling, with Measure Type "Number."

Time frame: Baseline up to week 120

Scale Factor Estimate for Final Parametric Time to Event PK/PD Model for PFS

Per planned analysis, Arms/Groups were combined \& PK/PD analysis for PFS was based on lenvatinib \& placebo data from this study combined with NCT01321554. Relationship between lenvatinib exposure \& PFS was assessed using Kaplan-Meier plots. A parametric survival model (proportional hazard model) with Weibull distribution structure was developed to estimate the probability distribution of time from study start to progression, as a function of covariates including baseline disease characteristics, demographics, lenvatinib exposure, changes in biomarker time profiles, model predicted change from baseline in tumor size \& change in tumor size time-profiles. Significant (p\<0.01) covariates from the univariate analysis were added to the model simultaneously \& significant predictors were retained according to backward exclusion criteria (log likelihood ratio test, p-value of 0.001). Data presented are scale factor estimated using non-linear mixed effects modeling, with Measure Type "Number."

Time frame: Time from the date of randomization to the date of first documentation of PD, or date of death, whichever occurred first up to approximately 3 years 3 months

Shape Factor Estimate for Final Parametric Time to Event PK/PD Model for PFS

Per planned analysis, Arms/Groups were combined \& PK/PD analysis for PFS was based on lenvatinib \& placebo data from this study combined with NCT01321554. Relationship between lenvatinib exposure \& PFS was assessed using Kaplan-Meier plots. A parametric survival model (proportional hazard model) with Weibull distribution structure was developed to estimate the probability distribution of time from study start to progression, as a function of covariates including baseline disease characteristics, demographics, lenvatinib exposure, changes in biomarker time profiles, model predicted change from baseline in tumor size \& change in tumor size time-profiles. Significant (p\<0.01) covariates from the univariate analysis were added to the model simultaneously \& significant predictors were retained according to backward exclusion criteria (log likelihood ratio test, p-value of 0.001). Data presented are shape factor estimated using non-linear mixed effects modeling, with Measure Type "Number."

Time frame: Time from the date of randomization to the date of first documentation of PD, or date of death, whichever occurred first up to approximately 3 years 3 months

Lenvatinib AUC Exposure Effect Estimate for Final Parametric Time to Event PK/PD Model for PFS

Per planned analysis, Arms/Groups were combined \& PK/PD analysis for PFS was based on lenvatinib \& placebo data from this study combined with NCT01321554. Relationship between lenvatinib exposure \& PFS was assessed using Kaplan-Meier plots. A parametric survival model (proportional hazard model) with Weibull distribution structure was developed to estimate probability distribution of time from study start to progression, as a function of covariates including baseline disease characteristics, demographics, lenvatinib exposure, changes in biomarker time profiles, model predicted change from baseline in tumor size \& change in tumor size time-profiles. Significant (p\<0.01) covariates from the univariate analysis were added to the model simultaneously \& significant predictors retained according to backward exclusion criteria (log likelihood ratio test, p-value of 0.001). Data presented are AUC exposure effect estimated using non-linear mixed effects modeling, with Measure Type "Number."

Time frame: Time from the date of randomization to the date of first documentation of PD, or date of death, whichever occurred first up to approximately 3 years 3 months

Predicted Percent Change in Tumor Size Estimate for Final Parametric Time to Event PK/PD Model for PFS

Per planned analysis, Arms/Groups were combined \& PK/PD analysis for PFS based on lenvatinib \& placebo data from this study combined with NCT01321554. Relationship between lenvatinib exposure \& PFS was assessed using Kaplan-Meier plots. A parametric survival model (proportional hazard model) with Weibull distribution structure was developed to estimate probability distribution of time from study start to progression, as a function of covariates including baseline disease characteristics, demographics, lenvatinib exposure, changes in biomarker time profiles, model predicted change from baseline tumor size \& change in tumor size time-profiles. Significant (p\<0.01) covariates from univariate analysis were added to the model simultaneously \& significant predictors retained according to backward exclusion criteria (log likelihood ratio test, p-value = 0.001). Data presented are the predicted change in tumor size estimated using non-linear mixed effects modeling, with Measure Type "Number."

Time frame: Time from the date of randomization to the date of first documentation of PD, or date of death, whichever occurred first up to approximately 3 years 3 months

Baseline Tumor Size Estimate for Final Parametric Time to Event PK/PD Model for PFS

Per planned analysis Arms/Groups were combined \& PK/PD analysis for PFS was based on lenvatinib \& placebo data from this study combined with E7080-G000-303(NCT01321554).Relationship between lenvatinib exposure \& PFS was assessed using Kaplan-Meier plots.Parametric survival model(proportional hazard model)with Weibull distribution structure was developed to estimate probability distribution of time from study start to progression as function of covariates including baseline disease characteristics,demographics,lenvatinib exposure,changes in biomarker time profiles,model predicted change from baseline tumor size \& change in tumor size time-profiles.Significant(p\<0.01)covariates from univariate analysis were added to model simultaneously \& significant predictors retained as per backward exclusion criteria(log likelihood ratio test,p-value of 0.001).Data presented are predicted baseline tumor size from the model with Measure Type "Number"estimated using non-linear mixed effects modelling.

Time frame: Time from the date of randomization to the date of first documentation of PD, or date of death, whichever occurred first up to approximately 3 years 3 months

Input Rate Indirect Effect Model Estimate From Base/Final PK/PD Blood Pressure Model

Per the planned analysis, Arms/Groups were combined and population PK/PD analysis for blood pressure was based on lenvatinib and placebo pooled data from this study combined with study E7080-G000-201 (NCT00784303) \& study E7080-G000-303 (NCT01321554). The effect of lenvatinib exposure (AUC) at the time of blood pressure assessment on systolic and diastolic blood pressure was tested as a simultaneous indirect model where lenvatinib AUC was linked to the input rate of the indirect effect model by a linear slope factor function. Based on the results from model development, an indirect PK/PD model with a linear effect of lenvatinib exposure on both systolic and diastolic blood pressure was selected as the base model for subsequent univariate analysis. The data presented are the input rate indirect effect model estimate, with Measure Type "Number." They are population PK/PD model predictions and have been estimated using mixed effects non-linear modeling.

Time frame: From date of first administration of study drug up to 6 months

Drug Effect on Systolic and Diastolic Input Rate Estimates From Base/Final PK/PD Blood Pressure Model

Time frame: From date of first administration of study drug up to 6 months

Number of Participants With Weight Decrease Stratified by AUC Quartile (Q) Group

Per the planned analysis for this endpoint, Arms/Groups were combined and a population PK/PD analysis of the relationship between lenvatinib exposure and occurrence of the TEAE weight decreased was based on placebo and lenvatinib pooled data from this study combined with study E7080-G000-201 (NCT00784303) and study E7080-G000-303 (NCT01321554). The relationship of occurrence probability of different grades of the TEAE weight decreased and lenvatinib exposure were evaluated by logistic regression model. The logit model was of the form: sum of intercept, of lenvatinib exposure, effects of covariates were explored, and random effects were used to describe between participant variability. Lenvatinib exposure was AUC based on the dose at the time of event. For each TEAE, probabilities of having no TEAE and a CTCAE Version 4.03 Grade 1 (Mild), Grade 2 (Moderate) or Grade 3 (severe or medically significant) TEAE were estimated as a function of lenvatinib or exposure.