Sorafenib as Adjuvant Treatment in the Prevention Of Recurrence of Hepatocellular Carcinoma (STORM)

Disease recurrence of HCC (intra or extra hepatic) was defined as the appearance of a new intrahepatic lesions fulfilling the American Association for the Study of Liver Diseases (AASLD) criteria of diagnosis of HCC or a new extra-hepatic lesions according to the Response Evaluation Criteria in Solid Tumors (RECIST) criteria version 1.0. In addition to investigator assessment, all images were reviewed by an independent panel of radiologists. The calculation of the RFS was based on the independent evaluation of the scans. RFS was defined as the time from randomization to the first documented disease recurrence by independent radiological assessment or death due to any cause whichever occurred first. For subjects who had not recurred or died at the time of analysis, RFS was censored at their last date of evaluable scan before drop-out for any other reason than recurrence or death.

Time frame: From randomization up to 4 years or until disease recurrence whichever came first

Population: Full analysis set (FAS) included participants who were randomized to study treatments.

OS was defined as the time from randomization to date of death due to any cause. OS for subjects alive at the time of analysis was censored at their last date of contact. NA in the reported data indicates values could not be estimated due to censored data.

Time frame: From randomization of the first subject until 4 years later.

Population: FAS included participants who were randomized to study treatments.

TTR was defined as the time from randomization to the first documented disease recurrence by independent radiological assessment. For subjects who had not recurred at the time of analysis, TTR was censored at their last date of evaluable scan before withdrawal for any other reason than recurrence. NA in the reported data indicates values could not be estimated due to censored data.

Time frame: From randomization up to 4 years or until disease recurrence whichever came first

Population: FAS included participants who were randomized to study treatments.

The EQ-5D is a generic quality of life preference based on a validated instrument used in cancer and in general population, with 2 parts: Index and Visual Analogue Scale. The EQ-5D Index is a descriptive system of the following health dimensions: mobility, selfcare, usual activities, pain/discomfort, and anxiety/depression. Subjects were asked to choose any one of the 3 response levels for each dimension: no problems, some problems, and severe problems. The 5 health dimensions were summarized into a single score, the EQ-5D Index score which ranged from -0.59 to 1 with higher scores representing better health states (0=death, 1= perfect health, and -0.59=a health state worse than death). A change of at least 0.10 to 0.12 points was considered a minimally important difference using Eastern Cooperative Oncology Group Performance Status as the anchor. The results on the Analysis of covariance of timeadjusted Area under curve for the EQ-5D index score were reported.

Time frame: Cycle (C) Day (D)1, C2D1, C3D1 and subsequent cycles up to C18, end of intervention visit

Population: FAS included participants who were randomized to study treatments.

The EQ-5D is a generic quality of life preference based on a validated instrument used in cancer and in general population, with 2 parts: Index and Visual Analogue Scale. The EQ-5D VAS is a measure that represents health status as a single value. It is a 20-centimetre vertical graduated visual analogue scale with scores that ranged from 0 (worst imaginable health state) to 100 (best imaginable health state). The respondent rated his/her current health state by drawing a line from the box marked 'your own health state today' to the appropriate point on the EQ-5D VAS. A 3-digit number (including leading zeros) was read off the scale from the point where the respondent's line crossed the scale, which was the EQ-5D VAS score. A change of at least 7 points on the VAS was considered as minimally important. The results on the ANCOVA analysis of time-adjusted AUC for the EQ-5D VAS score were reported.

Time frame: Cycle (C) Day (D)1, C2D1, C3D1 and subsequent cycles up to C18, end of intervention visit

Population: FAS included participants who were randomized to study treatments.

The PWB, FWB, SWB and EWB were summed to form the FACT-G total score. Subjects responded to each item on a 5-point Likert-type scale ranging from 0 (not at all) to 4 (very much). FACT-G scores ranged from 0 to 108 and the higher scores represented a better quality of life. The MID for the FACT-G total score was in the range of 6 to 7. The results on the ANCOVA analysis of time-adjusted AUC for the FACT-G score were reported.

Time frame: Cycle (C) Day (D)1, C2D1, C3D1 and subsequent cycles up to C18, end of intervention visit

Population: FAS included participants who were randomized to study treatments.

The FACT-HEP is a 45 item, self-administered, multi-dimensional, psychometrically sound questionnaire used extensively in oncology clinical trials. FACT-HEP consisted of five subscales: Physical Well-Being (PWB), Social Well-Being (SWB), Emotional Well-Being (EWB), Functional Well-Being (FWB), and Hepatobiliary Cancer Subscale (HCS). The PWB, FWB, SWB and EWB were summed to form the FACTGeneral (FACT-G) total score. The FACT-G and HCS scores were summed to form the FACT-HEP total score. FACT-HEP scores ranged from 0 to 180 and the higher scores represented a better quality of life. Subjects responded to each item on a 5-point Likert-type scale ranging from 0 (not at all) to 4 (very much). The minimally important difference (MID) for the FACT-Hep total score was in the range of 8 to 9. The results on the ANCOVA analysis of time-adjusted AUC for the FACT-HEP score were reported.

Time frame: Cycle (C) Day (D)1, C2D1, C3D1 and subsequent cycles up to C18, end of intervention visit

Population: FAS included participants who were randomized to study treatments.

Biomarker was analyzed at baseline \[i.e., before treatment\] as a dichotomized variable based on median biomarker levels, and dichotomized into high and low groups using an optimal max chi cut-off approach - not per intervention. As such, results were analyzed according to this stratification. Max-chi square methodology was used to search for the optimal cut point for dichotomization of each plasma biomarker and Kaplan-Meier curves were generated using the optimal cut point for each possible association examined. These biomarker analyses were retrospective and exploratory and of signal generating nature only.

Time frame: At Baseline

Population: Plasma biomarker analysis was done in 858 (77%) of the all enrolled subjects.

Biomarker was analyzed at baseline \[i.e., before treatment\] as a dichotomized variable based on median biomarker levels, and dichotomized into high and low groups using an optimal max chi cut-off approach - not per intervention. As such, results were analyzed according to this stratification. Max-chi square methodology was used to search for the optimal cut point for dichotomization of each plasma biomarker and Kaplan-Meier curves were generated using the optimal cut point for each possible association examined. These biomarker analyses were retrospective and exploratory and of signal generating nature only.

Time frame: At Baseline

Population: Plasma biomarker analysis was done in 858 (77%) of the all enrolled subjects.

Biomarker was analyzed at baseline \[i.e., before treatment\] as a dichotomized variable based on median biomarker levels, and dichotomized into high and low groups using an optimal max chi cut-off approach - not per intervention. As such, results were analyzed according to this stratification. Max-chi square methodology was used to search for the optimal cut point for dichotomization of each plasma biomarker and Kaplan-Meier curves were generated using the optimal cut point for each possible association examined. These biomarker analyses were retrospective and exploratory and of signal generating nature only.

Time frame: At Baseline

Population: Plasma biomarker analysis was done in 858 (77%) of the all enrolled subjects.