Table 2 Eastern Asian Panel's Opinions on Clinical Trial Design Aspects

Patient Population

Diagnosis

• Agree with AASLD recommendations[7] - pathological confirmation OR noninvasive criteria per AASLD guidelines

Target population

• BCLC stage is acceptable, but clinical protocols must account for portal vein involvement and liver function

• Treatment options for CP B/C are needed; CP B/C (ECOG PS 0 only) is an ideal population to study in advanced/metastatic HCC

Liver function

• Agree with AASLD recommendations[7]; however, trials should separately include and/or evaluate patients based on presence of cirrhosis or liver function grade.

Stratification

• Stratification by viral etiology is important in trials conducted within Eastern Asia

• Stratification by use of antivirals should also be considered

• Protocols should standardize antiviral therapy and include appropriate monitoring parameters

Treatment

Control arm for RCTs

• Heterogeneity in TACE/TAE practices must be addressed

• Placebo-controlled trials are feasible in unresectable disease, especially for those in whom locoregional therapy is indicated, pending maturity of post-TACE sorafenib data

• AASLD recommendation for sorafenib as comparator in advanced disease [7] is not necessarily reflective of real-world use in Eastern Asia at this time due to high cost and intolerable side effects

Phase-specific Clinical Trial Recommendations

Phase I

• Consider conducting Asia-specific phase I trials due to the potential for PK/PD differences between Asian and Western populations; however, Asian phase I trials may not be necessary for all targeted agents

• Population

• CP-A or CP score up to 7-8 (subgroup of CP-B) would be feasible for standard phase I trials

• CP-B with score 8-9 and CP-C could be enrolled in phase I trials testing agents at lower doses

Phase II

• For first-line studies in advanced HCC, AASLD recommendation for sorafenib [7] is not necessarily reflective of real-world use in Eastern Asia at this time due to high cost and intolerable side effects

• Agents demonstrated effective for second-line use in phase II trials (not necessarily phase III trials) can be compared to sorafenib in first-line studies

Phase III

• OS endpoint will soon no longer be appropriate in advanced disease with the introduction of multiple lines of therapies; PFS may be a surrogate but it is necessary to evaluate correlation with OS (ie, as what was done in colorectal cancer)

• In unresectable disease, the most appropriate endpoint is unknown due to difficulty distinguishing recurrence from second primary in the liver and unreliability of RECIST; time to development of new lesion is a possible endpoint

• Non-inferiority trials are acceptable if new agents have potential for less toxicity