Regorafenib as Second-Line Systemic Therapy May Change the Treatment Strategy and Management Paradigm for Hepatocellular Carcinoma

At the European Society of Medical Oncology World Congress of Gastrointestinal Cancer held in Barcelona, Spain, on 30th June 2016, positive outcomes were reported by the Study of Regorafenib after Sorafenib in Patients with Hepatocellular Carcinoma (RESORCE) trial, which investigated the efficacy of regorafenib as second-line therapy after sorafenib failure [1]. In this clinical trial, the group who received regorafenib achieved a survival benefit of approximately 2.8 months compared to the placebo group. Overall survival (OS) was 10.6 months in the regorafenib arm compared with 7.8 months in the placebo arm, with a hazard ratio (HR) of 0.62 (95% confidence interval [CI]: 0.50–0.78; p<0.001). These are groundbreaking results. The positive outcome achieved by this second-line systemic therapy is a major development, especially after the numerous reports of failures in clinical studies of first-and second-line systemic therapeutic agents (table ​(table1).1). Regorafenib therapy is expected to significantly prolong life expectancy by approximately 2.8 months in patients with hepatocellular carcinoma (HCC) who develop progressive disease (PD) during sorafenib therapy. This development will certainly lead to drastic changes in the treatment strategy and management paradigm for HCC. Table 1 Phase III Clinical Trials of Japanese Participation for HCC Design of the RESORCE Trial The RESORCE trial enrolled 573 patients with advanced HCC corresponding to Barcelona Clinic Liver Cancer (BCLC) stage B or C who were unresponsive to sorafenib. The patients were divided into placebo and regorafenib arms at a 1:2 ratio for the daily administration of placebo and oral regorafenib (160 mg), respectively, for three weeks on and one week off (four weeks/cycle) (fig. ​(fig.1).1). Geographic region, performance status on the Eastern Cooperative Oncology Group scale, α-fetoprotein level (≥400 or <400 ng/mL), macrovascular invasion, and extrahepatic disease were used as allocation factors. This study excluded patients who were intolerant of sorafenib and who discontinued the treatment because of side effects. It enrolled only those patients who discontinued sorafenib because of evidence of PD on imaging studies. In addition, patients were included only if they had received ≥400 mg sorafenib for at least 20 of 28 days immediately prior to radiologically detected PD. In other words, this trial was designed (1) to ensure regorafenib tolerance among patients, and to reduce the occurrence of the drug-specific skin symptoms because the compound is structurally similar to sorafenib [2,3] (fig. ​(fig.2)2) and (2) to reduce the effect of post-trial treatment on OS in both the placebo and treatment arms by using a homogeneous group of patients who developed PD due to sorafenib failure. Fig. 1 Design of the RESORCE Trial. ECOG PS=Eastern Cooperative Oncology Group Performance Status; RECIST=Response Evaluation Criteria in Solid Tumors. Fig. 2 Chemical structure of Regorafenib is very similar to that of Sorafenib. In general, post-progression survival (PPS) is defined as the time interval between the diagnosis of PD after primary treatment and the patient's death, and OS is the sum of PPS and progression-free survival (PFS). Therefore, even significant differences in PFS can be canceled out because PPS is prolonged. Indeed, OS showed a stronger correlation with PPS than with PFS in a clinical trial of sorafenib [4]. Because HCC responds extremely well to locoregional therapy, it is often used as post-trial treatment even in cases in which locoregional therapy is no longer applicable and molecular targeted agents are subsequently administered in accordance with the protocol, provided that the patient's general condition is stable. This rarely happens with other types of cancer and is therefore essentially unique to HCC, owing to the availability of powerful locoregional therapies such as intra-arterial infusion chemotherapy [5, 6, 7], transcatheter arterial chemoembolization (TACE) [8, 9], and radiofrequency ablation [10, 11, 12]. These post-trial treatments are capable of canceling out any difference in the primary endpoint OS by prolonging PPS [13]. Indeed, previous clinical trials of second-line agents other than regorafenib have always included patients intolerant to sorafenib, which may have increased the influence of post-trial treatment and thus contributed to their negative outcomes. Patients unresponsive to sorafenib are those who develop PD during sorafenib therapy and are likely to have relatively poor hepatic function and overall general condition. By contrast, patients intolerant to sorafenib are those who discontinue the treatment because of side effects; these patients are in relatively stable conditions because of negligible amounts of internalized sorafenib, and a lack of HCC progression. Because of their clinical stability, patients intolerant to sorafenib are inevitably treated by locoregional therapy or various other post-trial treatments, including the re-administration of sorafenib, regardless of whether they received an actual second-line agent or placebo during the trial. With this in mind, clinical trials of second-line agents should enroll only patients who are unresponsive to sorafenib [14]. The RESORCE trial was the first clinical study to reflect this point in the trial design (fig. ​(fig.1).1). The benefit of excluding patients intolerant to sorafenib was demonstrated in the subanalysis of a previous phase II study of axitinib, which generated an excellent HR and a significant study outcome [15, 16]. The second noteworthy point in the design of the RESORCE trial is that the allocation factors of macrovascular invasion and extrahepatic disease were treated as independent stratification factors. In general, the designs of previous clinical trials of molecular targeted agents involved allocation factors specifying “vascular invasion and/or extrahepatic spread” or “neither.” However, because vascular invasion is an extremely poor prognostic factor for HCC, assigning vascular invasion to the same category as extrahepatic spread may have influenced the outcome of these clinical trials. For example, when the treatment group contains more patients with vascular invasion but the placebo group includes more patients with extrahepatic spread, such sampling bias will put the treatment group at a significant disadvantage. In fact, such allocation imbalance apparently contributed to a negative outcome in a clinical trial of brivanib as second-line therapy [17] (table ​(table22). Table 2 Imbalance between Brivanib and Placebo Arm in BRISK-PS Trial The design of the RESORCE trial is excellent because it reflects what was learned from the negative outcomes of past trials and the reasons for those outcomes.