No established regimens exist for the treatment of patients with advanced hepatocellular carcinoma after failure of first-line treatment with the multikinase inhibitor sorafenib. Outcomes in the setting of sorafenib resistance or intolerance are poor, with a median expected overall survival for the placebo arms of second-line trials in the range of 7 to 8 months.
To date, no established regimens exist for the treatment of patients with advanced hepatocellular carcinoma (HCC) after failure of first-line treatment with the multikinase inhibitor sorafenib. Outcomes in the setting of sorafenib resistance or intolerance are poor, with a median expected overall survival (OS) for the placebo arms of second-line trials in the range of 7 to 8 months.
A recent study evaluating survival among 260 patients with HCC permanently discontinuing sorafenib reported median OS durations of 4.1 months overall, 4.6 months in patients with progression, and 1.8 months in patients with decompensated disease.1
The search for an active second-line regimen in advanced HCC has been marked by multiple disillusions: Within the past few years, 3 candidate agents failed to meet primary endpoints in large phase III trials, including brivanib (BRISK-PS study), everolimus (EVOLVE-1 study), and ramucirumab (REACH study).
Brivanib, a dual inhibitor of the vascular endothelial growth factor receptor (VEGFR) and fibroblast growth factor receptor (FGFR)mediated pathways, was the first agent evaluated in second-line therapy for HCC. A phase II study reported promising antitumor activity and a median OS of 9.79 months among patients with HCC resistant to antiangiogenic regimens, suggesting that FGF targeting may overcome resistance to antiangiogenic therapy.2However, the subsequent phase III BRISK-PS study involving 395 patients with HCC who progressed during or after sorafenib treatment or were intolerant to the drug failed to demonstrate a survival benefit versus placebo.3
Compared with placebo, brivanib was associated with significant improvement in objective response rate (ORR; 10% vs 2% with placebo;P= .003) and other secondary endpoints, including time to progression (TTP; 4.2 vs 2.7 months;P=.001) and disease control rate (DCR).3However, brivanib did not meet the prespecified primary endpoint of OS improvement (9.4 months vs 8.2 months with placebo; hazard ratio [HR], 0.89;P=.3307). The minimum HR of 0.78 to demonstrate positivity was also not met after adjustment for prognostic factors (adjusted HR, 0.81). Treatment-related toxicities were substantial, with a discontinuation rate secondary to adverse events (AEs) of 23%.3
Despite preliminary antitumor activity observed in phase I/II trials, the mTOR inhibitor everolimus failed to demonstrate clinical benefit in the phase III EVOLVE-1 study.4The survival of patients with HCC progression on or after sorafenib allocated to everolimus (n=362) or placebo (n=184) did not differ (median OS, 7.6 vs 7.3 months; HR, 1.05;P=.68); everolimus also did not significantly affect secondary endpoints including TTP (HR, 0.93).4
Most recently, the phase III REACH trial5of ramucirumab, a fully human IgG1 monoclonal antibody (mAb) targeting VEGFR-2, failed to detect a significant overall OS benefit versus placebo (median OS, 9.2 months vs 7.6 months; HR, 0.87;P=.14). Ramucirumab was associated with significant improvement in progression-free survival (PFS; 2.8 months vs 2.1 months; HR, 0.63;P<.001), ORR (7% vs <1%;P<.0001), and DCR (56% vs 46%;P=.011).5
Despite overall disappointing results, subgroup analysis of this trial revealed that patients with elevated baseline levels of serum alpha-fetoprotein (AFP), a marker of poor prognosis, benefited from ramucirumab, a finding that is being further investigated in an ongoing phase III trial.
Proposed reasons to explain the recent trial failures in second-line therapies include bias between treatment arms caused by inadequate patient stratification. According to David Bolos, MD, and Richard S. Finn, MD, from the David Geffen School of Medicine, University of California, Los Angeles, “strata used in the front-line setting may not be applicable in the second-line setting.”6In particular, separate consideration of 2 key prognostic factors for survival, extrahepatic spread and vascular invasion, has been proposed future trials.7
Other relevant factors to explain the recent phase III trial failures include insufficient exploration of liver toxicities in phase II trials that may have led to more discontinuations than anticipated; misinterpretation of the predictive value of the phase II trial endpoint TTP on OS; and lack of analysis of biomarkers that might predict response to treatment.8Consequently, the design of ongoing trials has been reconsidered and adapted to include more thorough exploration of toxicity and pharmacokinetics of novel agents in dose-escalation studies, as well as approaches for marker-guided patient selection.
Identifying Genetic Signature Personalized Therapy for HCC
The molecular heterogeneity of HCC, which may contain as many as 30 to 40 different mutations per tumor, including 5 to 8 mutations in putative oncogenic driver genes, renders the identification of key oncogenic pathways that would guide selection of targeted treatment challenging.
Encouraging findings for a potential enrichment strategy based on signaling pathway activation were made in a randomized phase II trial with the small-molecule MET inhibitor tivantinib: Patients with tumors expressing high levels of MET had a survival benefit, and tumor MET expression appeared both predictive and prognostic.9
The MET receptor tyrosine kinase and its ligand hepatocyte growth factor (HGF) are implied in carcinogenesis and metastatic progression in HCC; 25% to 87% of HCC overexpresses MET. High-MET expression is particularly frequent in advanced tumors with vascular invasion.
In the phase II trial, 107 patients with advanced HCC with progression on or intolerance to sorafenib were randomly allocated at a 2:1 ratio to tivantinib (n=71; 39 with initial starting dose of 360 mg twice daily, amended to 240 mg twice daily due the incidence of severe neutropenia) or placebo (n=36).9In the overall study population, tivantinib was associated with significantly longer TTP (1.6 vs 1.4 months; HR, 0.64; 90% CI, 0.43-0.94;P=.04).
Tumor MET expression was assessed by immunohistochemistry in 77 patients (72%), revealing high-MET tumors in 37 patients (48% of the study population). Patients with high-MET tumors had significant benefits with tivantinib treatment, resulting in prolongation of median TTP (2.7 months vs 1.4 months with placebo; HR, 0.43;P=.03) and of median OS (7.2 months vs 3.8 months with placebo; HR, 0.38;P=.01).9
Tumor MET status appeared predictive for response to treatment: Patients with low-MET tumors did not benefit from tivantinib treatment (median TTP, 1.5 months vs 1.4 months with placebo; HR, 0.96;P=.92; median OS, 5.0 months vs 9.0 months with placebo; HR, 1.33;P=.92).9
Outcomes from the phase II trial also suggest MET expression as a prognostic factor for survival in pretreated patients with HCC. Among patients receiving placebo, patients with high-MET tumors had a significantly shorter median OS than patients with low-MET tumors (3.8 months for high-MET vs 9.0 months for low-MET; HR, 2.94;P=.02).9
Tivantinib as second-line treatment in HCC is currently being investigated in a randomized phase III study in a selected population of patients with high-MET advanced HCC after sorafenib failure (METIV-HCC trial; ClinicalTrials.gov identifier NCT01755767). The primary endpoint is OS, and secondary endpoints include progression-free survival (PFS) and safety; the planned enrollment is 346 patients.
Promising outcomes in the second-line treatment of advanced HCC have also been observed with cabozantinib (XL-184), a small-molecule inhibitor of MET, RET, and VEGFR. A phase II study including 41 patients with advanced HCC reported clinical activity, with a 12-week DCR of 68%, median PFS of 4.4 months, and median OS of 15.1 months.10
Based on these results, a randomized phase III trial (CELESTIAL; ClinicalTrials.gov identifier NCT01908426) has been initiated to evaluate cabozantinib versus placebo in patients with advanced HCC that progressed following sorafenib. With a targeted recruitment of 760 subjects, the study projects 90% power to detect 31.6% increase in the primary endpoint, OS (HR, 0.76).
Biomarker-Driven and Novel Multi-Pathway Strategies
A different biomarker-driven approach has emerged from outcomes of the phase III REACH trial5evaluating ramucirumab. Although this study failed to detect a significant OS benefit in the overall study population, a prespecified subgroup of patients with elevated baseline AFP >400 ng/mL had a 20% ORR and significant improvement in survival after receiving ramucirumab (median OS of 7.8 months vs 4.2 months with placebo; HR, 0.67;P=.0059).5
The molecular mechanisms for this correlation remain unclear, but according to the authors, high α-fetoprotein expression may be associated “with a particular subtype of hepatocellular carcinoma with increased angiogenesis and possibly enhanced sensitivity to VEGFR-2 inhibition.”5
Based on these findings, the phase III REACH-2 trial (ClinicalTrials.gov identifier NCT02435433) has been initiated to assess ramucirumab in a second-line setting in patients with elevated AFP. In a post-hoc analysis of REACH, ramucirumab treatment led to a greater reduction in the risk of death in patients with progressively higher baseline AFP values.
The multistep oncogenic transformation underlying HCC involves alterations in several signaling cascades, resulting in molecularly heterogeneous tumors. Presumptive key carcinogenic signaling pathways include the Wnt/β-catenin pathway, chromatin remodeling, oxidative stress, and signaling involving EGF, PDGF, FGF, VEGF, and IGF, and intracellular mediators such as RAS/RAF/MAPK and PI3K/AKT. The simultaneous inhibition of multiple pathways is, therefore, an appealing concept that is currently being explored using the multikinase inhibitor regorafenib.
Structurally related to sorafenib, regorafenib is a potent inhibitor of kinases involved in angiogenic, oncogenic, and stromal/tumor environment-related signaling cascades, including VEGFR1-3 and TIE2, c-kit, Ret, BRAF, and PDGFR, among others.
Data from a single-arm phase II study have shown an acceptable safety profile and evidence of antitumor activity of regorafenib in patients with advanced HCC with progression on sorafenib.11Disease control was achieved in 26 of 36 patients (72%), median TTP was 4.3 months, and median OS was 13.8 months.11
Based on these outcomes, the phase III RESORCE trial (Regorafenib after Sorafenib in Patients with Hepatocellular Carcinoma; ClinicalTrials.gov identifier NCT01774344) has been initiated to compare the efficacy and tolerability of regorafenib versus placebo in patients with advanced HCC that has progressed on sorafenib. The primary endpoint is OS; secondary endpoints include TTP, PFS, objective tumor response, DCR, and safety.
Hope for Immunotherapy
Cancer immunotherapy has emerged as a highly effective therapeutic approach in multiple solid malignancies, and recent evidence suggests promise in the treatment of advanced HCC.
Data from an ongoing phase I/II trial of nivolumab, a fully human mAb directed against the immune checkpoint receptor programmed cell death protein-1 (PD-1) indicate safety and efficacy in patients with advanced HCC.13
Both the PD-1 and cytotoxic T lymphocyte-associated antigen-4 (CTLA- 4) inhibitory immune checkpoints normally temper T-cell responses to avoid damage to self; however, overexpression of these receptors and their ligands in cancers mediates escape from an antitumor response.
Nivolumab prevents interaction of PD-1 with its ligands, augmenting antitumor immunity. Overexpression of PD-L1 is observed in 45% to 95% of HCC and has been associated with a poor prognosis.
The dose-escalation part of the study is enrolling patients with advanced HCC with progressive disease after at least 1 line of systemic therapy. Of the 47 patients evaluable for safety at the interim analysis in March 2015, 70% had extrahepatic metastases or vascular invasion, 75% had received prior lines of systemic therapy that failed, and 68% had received prior therapy with sorafenib.13
Patients were treated in parallel cohorts based on presence or absence of active viral hepatitis, and received 0.1 to 10.0 mg/kg of nivolumab intravenously for up to 2 years; a maximum tolerated dose was not defined.
Of 39 patients evaluable for response, 2 had a complete response (CR; 5%) and 7 had partial response (PR; 18%); 18 patients had stable disease (SD). Responses were durable (CR, 14 to >17 months; PR, <1 to >8 months; SD, 1.5 to >17 months). The OS rate at 6 months was 72%, and preliminary 1-year OS was 62%.
Recruitment to the phase I study (ClinicalTrials.gov identifier NCT01658878) with an estimated enrollment of 600 patients is ongoing. The expansion phase of the study will include both patients with progression on sorafenib and patients with no prior exposure to sorafenib or intolerance to sorafenib.
Emerging preclinical data suggest that the combination of a stromal cellderived 1-alpha receptor (CXCR4) inhibitor and PD-1 immunotherapy may overcome HCC progression and resistance after sorafenib treatment. Exposure of HCC to sorafenib has been associated with intratumoral hypoxia that results in increased expression of regulatory proteins involved in immune evasion, such as PD-1, and of components of signaling pathways that promote angiogenesis, tumor progression, and metastasis. In orthotopic mouse models of HCC, the combination of CXCR4 and PD-1 inhibition with sorafenib produced increased tumor cell death and caused tumor regression in cirrhotic livers.14
“Blockade of both CXCR4 and PD-1 prevents suppression of immune cell function in HCC tumors, enhances immune cell tumor penetration and activation, and ultimately delays HCC progression,” the authors noted. “Further understanding of the complex interaction between antiangiogenic, antimetastatic, and immunotherapeutic agents is warranted to facilitate progress in the systemic treatment of advanced HCC.”14