Changed frequency of hematology testing and dose reduction requirements following recommendations from Data Monitoring Committee (DMC). Clarified in synopsis minimum requirement for regional lymph node to be considered involved (previously clarification was only in body of protocol). Corrected language that remained in error in the original protocol regarding stopping rules for subject treatment. Protocol now was consistent throughout with existing details in Section 4.2.1, Table 4.1. Updated to reflect the fact that the LabCorp IHC assay now has IDE approval for investigational use in this study and that assay has CE mark in EU. .Revised minimum number of tissue slides required per current lab manual. Revised language to account for the fact that some sites have a single consent form for the entire study while others had separate consent forms for various portions of the study. Corrected discrepancy in protocol regarding timing of PK on Cycle 1 Day 22. Protocol now was consistent throughout with existing details in Section 8.1. Following multiple queries, added clarification on use, storage and destruction of tissue samples for this study. Corrected typo (symbol for ≤ corrected to ≥) and added clarity on Hy’s Law definition.

Added exclusion criterion number 14 to exclude subjects with pleural effusion or clinically evident (visible or palpable) ascites following recommendations from Data Monitoring Committee (DMC). Such subjects may have been more susceptible to infections and more at risk if they were to develop neutropenia. Reduced starting dose from 240 mg BID to 120 mg BID following recommendations from DMC as higher than expected rate of severe neutropenia was seen at original starting dose of 240 mg BID and exposure was noted to be higherthan it was in the phase 2 HCC study. Updated clinical experience section and study rationale section to distinguish which prior studies used capsule formulation and which used tablet formulation and to clarify that in this study tablet formulation was used. Updated dose modifications section to allow one lower level dose reduction given that starting dose had been reduced following recommendations from DMC. Changed visit schedule to increase frequency of hematology testing during Cycle 1 following recommendations from DMC. Hematology must have now been performed every 2 days during cycle 1. This would help more closely monitor any changes in neutrophil values and take action more quickly if necessary. Updated contact information for CRO physician to generic contact as named physician was terminating employment. The updated contact information was valid even if the assigned CRO physician changed again in the future.

Added description and rationale for the change of study design (eg, reduced dose) and subject grouping (2 cohorts: 120 mg cohort and 240 mg cohort) for statistical analyses. Updated study design due to reduced dose. Updated synopsis and statistical methods section to reflect that the efficacy analyses was to be performed only for the 120 mg cohort as 120 mg BID was the intended dose regimen for approval. Updated statistical methods section to reflect that the safety analyses and summary of disposition, demographic and baseline characteristics, and exposure were to be performed separately for the 120 mg and 240 mg cohorts. Updated the section of Risks and Benefits for Study Subjects to incorporate additional DMC review for the 120 mg cohort based on the DMC recommendation. An administrative update to make protocol PK and ECG schedule consistent with section 6.5.2. Editorial changes

Updated to confirm enrollment could continue per new recommendations from the DMC after complete review of neutrophil count data on 15 November, 2013. For consistency, updated protocol so that retesting of hematology through resolution of AE was consistent regardless of whether neutropenia was grade 2, 3 or 4. Reduced frequency of hematology monitoring during the first cycle. Of the 29 subjects treated at 120 mg BID by 15 November 2013, 26 were treated from between 1 and 2½ months and 3 were treated for at least 3 weeks. After reviewing the Absolute Neutrophil Count (ANC) data provided on 15 November 2013, the DMC advised that the frequency of hematology monitoring during the first cycle could be reduced from every 2 days to every 4 days. Re-wrote sections regarding re-screening of subjects to more clearly distinguish between subjects requiring full re-screening and those who needed to send in fresh biopsy after MET low result initially received. Protocol section 4.2.3 and 6.1 were now consistent with footnote 6 in Appendix 17.1. Furthermore we clarified the timeframe between notification of MET-high status and initiation of screening procedures. Corrected typographical error in Section 5.1.2. Following suggestions from some regulatory agencies, modified the frequency of hematology testing post Cycle 1 to allow for a more gradual decrease in frequency of hematology testing between Cycle 1 and subsequent cycles. Instead of testing hematology twice a month in Cycle 2, it would be tested on weekly basis during Cycle 2 before reverting to twice a month testing in subsequent cycles.

Updated current address of ArQule, Inc. co-sponsor of the study. Changed sponsor representative who signs the protocol approval page. In the event of a positive study showing significant difference favoring tivantinib over placebo, updated various sections in the protocol to allow for the possibility of subjects randomized to placebo to receive active treatment with tivantinib. This would occur after database lock and unblinding. The protocol was amended also to specify the safety eligibility criteria such patients would have to meet in order to receive active treatment. Added a definition of “end of trial” to the protocol. Added a sentence to clarify that after database lock, additional data will be reviewed via listings. Clarified that subjects unblinded before database lock would have to discontinue treatment immediately. Added updated standardized language to the protocol regarding drug accountability (which matches the process being followed to-date). Removed language requiring that MRI be used exclusively in certain countries. Listed safety procedures that are required at the time they start tivantinib for any subjects randomized to placebo who then receives tivantinib after database lock and unblinding (if the study is positive). Corrected typo to clarify that Alpha Fetoprotein (AFP) is collected at screening and not on Day 1. This is now consistent throughout all sections of the protocol. Added clarification language to one sentence in the statistical section. Updated wording in Child Pugh appendix to make consistent with wording in the body of the protocol.