Eastern asian expert panel opinion: designing clinical trials of molecular targeted therapy for hepatocellular carcinoma

Over 600,000 cases of hepatocellular carcinoma (HCC) are diagnosed annually worldwide and the mortality-to-incidence rate ratio is second only to pancreatic cancer [1, 2]. The incidence of HCC varies widely by geographical region. Asia carries the largest burden with 55% of all cases occurring in China [1]. Age-standardized incidence rates per 100,000 persons for men are 45.0 in Korea (1999-2001) [3], 37.9 in China (2002) [1], and 23.1 in Japan (2002) [1]. Corresponding rates for women are 12.0, 14.2, and 7.6. Globally, the predominant cause of HCC is viral infection with hepatitis B virus (HBV) or hepatitis C virus (HCV) [4].

Hepatocellular carcinoma is refractory to cytotoxic chemotherapy [5] and the failure of cytotoxic regimens has led to a bleak outlook. However, the recent development of molecular targeted therapies is changing the landscape and offering hope. Researchers have found new optimism for initiating clinical trials after sorafenib showed efficacy in advanced disease [6]. Currently, trials are planned or ongoing in all stages of HCC; however, many issues remain [7]. Most salient is the variability in management practices both between Asia and the West and within Asia. Key differences are apparent in the etiology, diagnosis, staging, and treatment of HCC among countries. These differences complicate the conduct of international clinical trials that will foster approval and availability of new therapeutic entities.

In order to forge a better understanding of how HCC clinical practices in the Eastern Asian region compare to current global clinical trial requirements, an expert conference was held. Participants of the panel (the authors) are oncologists and hepatologists representing China, Hong Kong, Japan, Korea, and Taiwan who have an expertise in treating HCC. Each panelist offered insight, reviewed herein, about how HCC is managed across Eastern Asia and how management practices and clinical trial requirements can be unified to advance new treatments, particularly targeted agents, for HCC.

Viral etiology varies by region with HBV predominating in non-Japanese Asians and accounting for approximately 70-80% of cases. In Japan, most of Europe, and in the United States, HCV is more common than HBV among viral etiologies [3, 8‐11]. However, in the United States, 67% of HCC cases are seronegative for both viruses [10].

The increased incidence of HBV-HCC in Eastern Asia compared to Japan and Western nations leads to different management issues and prognosis that affect clinical trial design. Hepatitis C virus-HCC is more likely to develop in the background of cirrhosis than HBV-HCC [12]. Therefore, the underlying liver disease may differ in HCC patients by region, a factor that weighs heavily in treatment decisions.

Survival differences have been observed according to geographic region and viral etiology, though the reasons for these observations remain unclear. In clinical trials of systemic therapy for advanced HCC, trials done in Asian countries reported inferior survival compared with trials done in non-Asian countries [13]. Possible reasons include variation in genetic and/or epigenetic aberrations between different viral etiologies and the propensity for Asian physicians to use local therapy more aggressively and in later stages, resulting in enrollment of a more advanced patient population to trials of systemic therapy. Survival between HBV-HCC and HCV-HCC appears similar in early-stage, resectable HCC, if staging and other clinical parameters are considered [14]. However, two retrospective studies have found poorer survival in HBV-HCC among patients with unresectable, advanced disease [15, 16]. Attributing the survival difference to viral etiology alone is difficult but demonstrates the need for considering the potential differences in clinical trials.

Additionally, in contrast to HCV, HBV reactivates with immune suppression, complicating treatment with immunosuppressive regimens [17, 18]. The predominance of HBV-HCC in Asia is associated with increased use of antiviral agents to prevent viral reactivation during HCC treatment. Antiviral therapy with lamivudine has reduced the incidences of HBV reactivation and hepatitis, reduced the severity of hepatitis episodes, led to fewer disruptions in chemotherapy, and reduced mortality related to HBV reactivation in clinical trials of patients with HCC or other cancers who are receiving chemotherapy [19‐22]. Anti-viral therapy following curative resection, radiofrequency ablation, or other local, non-chemotherapeutic treatments for HBV-HCC, has been shown to increase residual liver volume and/or function and may prolong survival [23‐25]. Furthermore, interferon, given after curative therapy, may increase recurrence-free survival rates [26, 27]. These benefits indicate that use of antiviral therapy is an important confounding factor in HCC clinical trials.

A separate international expert panel has recommended stratification according to region for global trials but discouraged further stratification according to etiology [7]. However, in light of the confounding factors described herein, the current panel agreed that trials within Eastern Asia should include stratification by HBV or HCV etiology. Further, antiviral therapy should be both considered as a stratification factor and incorporated into the overall management of patients in international HCC clinical trials.

Stage at diagnosis differs both within Eastern Asia and between Eastern Asia and Western nations. Using TNM-based staging systems, China and Japan have relatively high proportion of patients diagnosed at Stage I or II compared to Hong Kong and Korea. In the United States, a higher percentage of patients are diagnosed with distant metastasis compared to Asian countries [28, 29].

The differences may reflect variable screening practices. The proportion of patients who receive screening in the United States appears to vary according to the individual's healthcare. Only 25% of family practice physicians report routinely screening appropriate patients for HCC compared to 84% of physicians who are members of the Association for the Study of Liver Diseases (AASLD) [30, 31]. In a study of 157 patients diagnosed with HCC at three US Veteran Affairs (VA) medical centers, 39% of patients with a known risk factor for HCC received screening [32]. With the exception of Hong Kong, where screening has been conducted in the context of study, screening high-risk populations is the standard of care in Asia. With diagnosis occurring at earlier stages, Eastern Asian countries are better able to utilize curative therapies, significantly affecting treatment paradigms and clinical trial populations.

Both pathological and clinical diagnostic procedures vary according to country. The majority of pathological diagnoses are made by core biopsy in Korea, China, and Hong Kong, with fine needle aspiration (FNA) used infrequently. In contrast, 30% or fewer of pathological diagnoses are made by core biopsy in Japan, and Taiwan. Taiwan employs FNA in approximately 10% of cases but utilizes surgery for pathologic diagnosis in approximately 38% of cases. Protocols designating biopsy-proven HCC as an enrollment requirement would conflict with current practices in Japan and Taiwan. The panel agreed that for trials conducted in the advanced/metastatic setting, histological confirmation of HCC is not necessary. Further, pre-treatment biopsy may result in tumor seeding which would complicate neoadjuvant trials.

A variety of staging systems are employed worldwide [33‐36]. Several of these systems are based on the tumor-node-metastasis (TNM) paradigm or incorporate TNM groupings as a variable [33‐35, 37]. Other systems, such as the Barcelona-Clinic Liver Cancer (BCLC) staging system, incorporate measures of liver function and underlying disease. Complicating international clinical trial design is the variable use of these systems both within Asia and globally. Each region of Asia represented by the panel currently utilizes a different system. In China, the revised Staging Criteria of Primary Liver Cancer is used. This system was developed by the Chinese Society of Liver Cancer. The system uses criteria based on size, number and location of tumors, lymph node spread, extrahepatic metastasis, portal vein thrombosis, and liver function (Child-Pugh scores) [38]. In Japan, both the staging system and treatment algorithm apply liver function as the first category of evaluation rather than tumor size. Hong Kong does not have a unified staging system. Although BCLC is considered a valuable tool for a treatment algorithm in Hong Kong, the system is considered less useful for prognostication. The Chinese University Prognostic Index (CUPI) [37] has been found useful for prognostication at one center due to the more advanced population [39]. Korea employs a modified International Union Against Cancer (UICC) system and Taiwan uses BCLC.

The TNM-based staging systems have an important drawback: these systems do not account for underlying liver disease [40]. In HCC, the presence of liver disease is a common and important prognostic factor that is integral in determining treatment [40, 41]. For these reasons, TNM-based systems have limited value in the comprehensive management of HCC. The Child-Pugh (CP) score is a widely-accepted system to evaluate liver function. Despite empirical selection of variables, this tool represents a simple, bedside tool that predicts mortality in cirrhotic patients with a degree of accuracy not substantially less than the more statistically sound model for end-stage liver disease (MELD) [42]. The BCLC staging system incorporates measures of liver function (portal hypertension, bilirubin, and CP scores at higher stages) and has emerged as the standard for clinical trial design [6, 43]. However, this system is not generally used in Eastern Asia with the exception of Taiwan. China, specifically, has failed to adopt this system due to the omission of portal vein thrombosis as a factor, which has been shown to independently predict mortality [41]. Additionally, BCLC includes portal venous hypertension which requires an invasive procedure to measure that is not standard practice in Asia. However, the panel indicated that, if required for clinical trials seeking United States Food and Drug Administration approval, BCLC would be acceptable if the protocols also incorporated portal vein hypertension - measured and defined with non-invasive standardized methodology - and further evaluation of liver function.

Treatment practices vary somewhat throughout Eastern Asia and no unified treatment algorithm exists. Japan, China, Hong Kong, Korea, and Taiwan each use separate treatment algorithms, all of which differ from the BCLC treatment algorithm [7, 44, 45]. Such variations in treatment practices cause challenges in defining treatment protocols for international clinical trials.

In 2008, the American Association for the Study of Liver Diseases (AASLD) published a framework for clinical trial design in HCC [7]. During the current expert panel meeting, participants provided their views about clinical trial design from an Asian perspective. These views are outlined in Table 2.

The Asian panel also provided additional insights into clinical trial issues specific to disease stage. The panel noted a great need for trials in resectable disease. The panel felt that testing compounds in the adjuvant setting before establishing efficacy in the metastatic setting is possible, citing positive phase II adjuvant results with muparfostat (formerly PI-88) [72] and noting the need for effective therapies in this setting. The panel also expressed interest in chemoprevention with sorafenib and other agents after resection or local ablation. In unresectable disease, especially where locoregional therapy is indicated, placebo-controlled trials remain feasible, though the panel acknowledged opportunities are limited. In this setting, it may be beneficial to limit enrollment to patients who experience a maximal response after TACE based on modified EASL criteria [73]. Such a requirement would facilitate identification of subsequent disease progression across patients. However, additional research is necessary to identify the best clinical endpoints in this setting. Because it remains difficult to differentiate recurrent disease from a second primary cancer, time to development of a new lesion may be an appropriate outcome in this setting. Finally, in the advanced/metastatic setting, the panel felt that developing new agents in the second-line setting is warranted.

Hepatocellular carcinoma is a disease of variable incidence and etiology that is managed differently worldwide. This expert panel has identified key areas that need to be addressed to facilitate clinical trials in Asia. Stratification by viral etiology is desirable within Asia and by region in global trials. Antiviral therapy should also be considered as a stratification factor and incorporated into HCC management in trials. The panel agreed with AASLD that histological diagnosis is not required for trial entry. Staging and treatment plans vary significantly. The panel felt BCLC staging is acceptable for trials as long as portal vein hypertension can be measured and defined with non-invasive standardized methodology and liver disease is further evaluated. Consensus in treatment must be sought to allow multi-national trials and it must be recognized that first-line sorafenib is not largely feasible in Asia. Finally, Asian nations must be urged to participate in clinical trials, many of which are ongoing, to advance new treatment options in this challenging disease.