The different drugs (available or under development) are classified according to their putative targets (Table 1): drugs targeting the signal transduction via blockade of one or more growth factor receptors and/or intracellular signalling, and agents targeting angiogenesis, apoptosis, cell cycle, cell migration or protein turnover. This classification is arbitrary as most drugs are not specific for only one target or pathway but overlap several. These targeted drugs are mainly ligands specific to cell surface receptors (antibodies) and protein kinase inhibitors.
Targeting the epidermal growth factor receptor/mitogen-activated protein kinases pathway
One of the best-characterized signalling pathways is the epidermal growth factor receptor (EGFR)/mitogen-activated protein kinases (MAPK) pathway, which includes receptor EGFR and a cascade of phosphorylations involving at least four kinases: Ras, Raf, mitogen-activated protein extracellular kinase (MEK) and extracellular signal-regulated kinase (ERK) (17). The ligands EGF, platelet-derived growth factor (PDGF), hepatocyte growth factor (HGF) and VEGF, among others, can activate the Ras/MAPK signalling pathway, which is a focal point for cell proliferation and differentiation signal transduction. Upregulated activity of the MAPK pathway has been well documented in HCC cell lines, in in vivo HCC models and in human HCC specimens (13). A recent study showed that the HCV core protein could directly activate the Raf/MEK/ERK pathway, raising the potential of a major role of this pathway in virus C-induced hepatocarcinogenesis (18). Loss of Raf kinase inhibitor protein promotes HCC proliferation and migration. Different approaches have been developed to inhibit the EGFR/Ras/Raf/MAPK pathway at different levels.
Antagonization EGFR function can be carried out using a neutralizing monoclonal antibody such as cetuximab and panitumumab, or using a tyrosine kinase inhibitor (TKI) of EGFR activity, such as gefitinib, erlotinib and lapatinib. Phase II trials with erlotinb (19), gefitinib (20), lapatinib (21) and cetuximab (22, 23) are available. In all of them, the stage of the patients is difficult to understand and hence, data about survival are not highly informative for deriving a valuable signal in terms of benefit. In most cases, the main end-point is progression-free survival (PFS) at 6 months (median PFS ranging from 1.7 to 4.7 months) and it is impossible to rule out a confounding effect of cirrhotic death in all the studies. Disease control rate (partial response+stable disease rates) ranges from 20 to 60%. The association of cetuximab with chemotherapy (gemcitabin, oxaliplatine and capecitabine) has reported a low response rate but a high rate of stable disease, leading to a time to disease progression of about 5 months (22). Results of a phase II trial with the erlotinib–bevacizumab association (a powerful anti-anagiogenic) are described later (24).
The inhibition of the Ras/Raf/MAPK pathway may have a huge potential.
(a) Targeting Ras with inhibitors of farnesyl transferase is of interest as overexpression of Ras has been demonstrated in HCC and dysplasic liver nodules. To be competent for signal transduction, Ras proteins require post-translational modification by incorporation of prenyl moieties (farnesyl and geranyl groups). One of these inhibitors, ABT-100 (prevents prenylation of Ras proteins), has shown efficacy in preventing the development of chemically induced HCC in rats.
(b) Targeting Raf with Raf kinase inhibitor Bay-439006 (sorafenib). Sorafenib is a multikinase inhibitor targeting the Raf serine/threonine kinases and the VEGFR 1/2/3, PDGFb, c-kit, Flt3 and p38 tyrosine kinases (25). In vitro, sorafenib has been shown to stop tumour cells growth and induce apoptosis of HCC cell lines through inhibition of Raf/MEK/ERK signalling (26). It has also been shown to inhibit the growth of human HCC xenografts in nude mice (27). Finally, sorafenib has successfully undergone phase I (28), II (29) and III trials (16, 30). The results of the main trials are summarized in Table 2 and Fig. 2. The median survival of patients is increased from 7.9 to 10.7 months (hazard ratio 0.69, P<0.0001). The positive data obtained in the phase 3 trial have been validated in a recent Asian trial (30) and hence, currently, there is no doubt that sorafenib is now the standard of care for patients with advanced HCC (15, 31). It increases the life expectancy in more than 40% of patients compared with placebo and has a safe profile in terms of side effects. All strata of patients and clinical profiles appear to benefit from sorafenib (Table 3 and Fig. 3). The magnitude of the benefit obtained by sorafenib is equal to that offered by targeted therapies in patients with advanced colorectal (32), lung (33), breast (34) and head and neck cancers (35), among others, both in terms of likelihood of death and in terms of absolute values. Hence, the treatment of advanced HCC deserves the same positive evaluation as that in other cancers.
Table 2. Randomized controlled trials of sorafenib in advanced hepatocellular carcinoma
| ||Llovet et al. (16)||Cheng et al. (30)|
|Regimen||Sorafenib 800 mg/day||Sorafenib 800 mg/day|
|Number of patients||299 treated (303 placebo)||150 treated (76 placebo)|
|Median age||65 years||51 years|
|Primary end point||Survival||Survival|
|Classification BCLC B/C(%)||18%/82%||4%/96%|
|Metastasis/vascular invasion/AFP||51%/36%/median 44 ng/ml||68%/34%/NA|
|Etiology of cirrhosis HCV/HBV/alcohol/other||29%/19%/26%/29%||11%/71%/18%|
|Response rate||DCR 43%||DCR 35%|
|TTP (control)||5.5 months (2.8 months)||2.8 months (1.4 months)|
|HR (95% CI)||0.58 (0.45–74)||0.57 (0.42–0.79)|
|Median survival (control)||10.7 months (7.9 months)||6.5 months (4.2 months)|
|HR (95% CI)||0.68 (0.55–0.87)||0.68 (0.50–0.93)|
|Toxicity grade 3/4||30%||24%|
Figure 2. Survival (a) and time to progression curves (b) of patients randomized to sorafenib or placebo. Adapted from Llovet (16).
Download figure to PowerPoint
Table 3. Sorafenib therapy in patients with advanced hepatocellular carcinoma, subgroup analysis of SHARP trial
| ||Alcohol related cirrhosis (58)||Presence of vascular invasion (MVI) and or extrahepatic spread (EHS) (59)||Absence (PS 0) or presence of symptoms (PS 1–2) (60)||Prior antitumour therapy||Baseline transaminases (61)||HCV cirrhosis (62)|
|Number of patients||79/80||MVI-EHS: 209/212||PS 0: 161/164||Prior curative treatment: 81/77||AST/ALT normal: 143/149||93/85|
|Sorafenib/placebo|| ||No MVI-EHS: 90/91||PS 1–2: 138/139||Prior TACE: 86/90||AST/ALT mild: 85/75|| |
| || || || ||AST/ALT moderate: 68/78|| |
|TTP median (months)||5.5/3.9||MVI-EHS: 4.1/2.7||PS 0: 5.5/2.9 (s)||Prior curative treatment: 5.5/2.7||AST/ALT normal: 5.5/3.3||7.5/2.7|
|Sorafenib/placebo|| ||No MVI-EHS: 9.6/4.3||PS 1–2: 5.3/2.8 (s)||Prior TACE: 5.8/4||AST/ALT mild: 5.5/2.8|| |
| || || || ||AST/ALT moderate 5.5/2.7|| |
|Median survival||10.3/7.9 months||MVI-EHS: 8.9/6.7 months||PS 0: 13.3/8.8 months||Prior curative treatment: 11.9/8.8||AST/ALT normal: 13/9||14/7.9|
|Sorafenib/placebo|| ||No MVI-EHS: 14.5/10.2 months||PS 1–2: 8.9/5.6 months||Prior TACE: 11.9/9.9||AST/ALT mild: 11/8|| |
| || || || ||AST/ALT moderate: 8/5.5|| |
It is important to stress that the impact of sorafenib has been achieved without a significant effect in terms of the response rate. A minority of patients experience tumour reduction according to the RECIST criteria (36), and the effect was captured when assessing time to progression (TTP). This was significantly delayed and shows that the conventional oncology definitions will have to be modified because the activity of some of the new agents is not captured by the usual tools merely based on size reduction (31).
The phase 3 trials mostly included patients in Child–Pugh A stage. Pharmacokinetic data are similar in patients in Child–Pugh B, and on comparison with the data of phase 2 investigations, it appears that the efficacy and safety is maintained (37). Obviously, Child–Pugh B is a very wide group in terms of liver function impairment and life expectancy. Thus, it appears that at Child–Pugh B 7 points, treatment effectiveness would be maintained, while data in more evolved liver disease are unknown, and indication of sorafenib should be based on an individualized medical evaluation.
It is very appealing to combine sorafenib with other agents, either chemotherapy and/or other targeted therapy. A recent review summarized the efficacy, safety and pharmacokinetics of sorafenib with other targeted agents or cytotoxic drugs from a series or phase I/II trials in 600 patients with various solid tumours (38). Common dose-limiting toxicities were the hand foot syndrome, diarrhoea and fatigue. The maximum tolerated dose (MTD) was not reached in most trials, except with bevacizumab association. Sorafenib had little effect on the pharmacokinetics of the co-administred agent, and vice versa, except with doxorubicin, irinotecan and docetaxel. These alterations were not associated with increased clinical toxicity. There are no robust data to establish the efficacy of the combinations, but clearly a careful approach is mandatory to not lose the benefit in survival because of major toxicity because of the combination.
A recent phase II randomized study included 96 patients treated with sorafenib plus doxorubicin or doxorubicin plus placebo. The response rate was weak but similar (4%). Results for TTP and OS favour the sorafenib–doxorubicin association vs. doxorubicin–placebo (TTP 8.6 vs. 4.8 months; OS 13.7 vs. 6.5 months). An increased risk of grade 3/4 neutropenia was noted in the association group (53 vs. 46%) (39). Obviously, a more convenient study would have been the comparison of sorafenib with the sorafenib/doxorubicin association, as, in its current format, the sole conclusion is that sorafenib retains its efficacy in the presence of doxorubicin, while there is no proof that doxorubicin is of any help.
(c) Targeting MEK kinase. A recent study has shown that a MEK inhibitor could stop HCC proliferation and growth of a human HCC xenograft (40). However, long-term administration of the MEK inhibitor led to the development of resistance and loss of drug efficacy.
Targeting the PI3K/Akt/mTOR pathway
The second well-defined pathway is the PI3K/AKT/mTOR. Akt can be activated by different ways: through a tyrosine kinase receptor [EGF or insulin growth factor (IGF)], through a constitutive activation of PI3K or through the loss of function of the tumour suppressor gene PTEN (mutational acquired). PI3K activates serine/threonine kinase Akt, causing FOXO3a phosphorylation, and its nuclear translocation, promoting apoptosis. There is a strong relationship between a low level of FOXO3a evaluated by reverse transcriptase polymerase chain reaction or immunoblotting and survival (a two-fold decrease in expression vs. normal) (41). An important mediator of this pathway is mammalian target of rapamycin (mTOR). mTOR is a central regulator of cell growth and proliferation and is activated in a subset of HCCs (42).
Mammalian target of rapamycin inhibitors: rapamycin and analogues everolimus and temsirolimus are used as immunosuppressive therapies after liver transplantation, and for advanced renal cell cancer (43). Treatment of HCC xenograft mice with mTOR inhibitors caused significant tumour shrinkage and a reduced level of FOXO3a (44).
Perifosine is a novel oral alkylphosphocholine with effects on multiple signal transduction pathways including Akt, MAPK and Jun. The drug was given to 13 patients with HCC at a daily dose of 50 mg: one partial response and four stable diseases over 6 months were noted with no severe side effects (45).
Angiogenesis is a fundamental process for tumour growth. Cells of the growing tumour tissue are exposed to physiological stresses connected with insufficient delivery of oxygen (hypoxia) and accumulation of acidic products of the glycolytic metabolism (acidosis). Adaptation to these micro-environmental stresses leads to an increased secretion of the hypoxia-inducible factor (HIF), which activates a broad array of genes functionally involved in angiogenesis such as VEGF and PDGF. Transmembrane carbonic anhydrase isoform IX (CA IX) is a direct target of HIF and serves as a surrogate marker of hypoxia. Tumour and stromal cells can also secrete VEGF; thus, VEGF acts as a stimulating growth factor for endothelial cells (paracrine stimulation), tumoral and stromal cells (autocrine stimulation). VEGF is also inducible via other pathways such as Ras/Raf/MAPK, Akt/PI3K/mTOR and PTEN. An increased level of VEGF and microvessel density is a common pattern in many cancers, and also in HCC. Targeting angiogenesis is of particular interest in HCC, because of its hypervascular profile.
Bevacizumab is a VEGF monoclonal antibody with some available data in HCC treatment as a single agent or in association. Bevacizumab seems to give more tumoral responses than sorafenib according to RECIST; however, the impact on survival remains unknown. Bevacizumab also has more contra-indications than EGFR inhibitors (vascular and cardiovascular), and has to be avoided in patients with vascular or renal impairment. In addition to these limitations, the data available on bevacizumab indicate a high risk of bleeding, in some cases leading to death (50).
TSU-68 is an oral multikinase inhibitor (VEGFR, PDGFR and FGFR). In a phase II study, TSU-68 was given at a daily dose of 400 mg to 35 Child-Pugh A/B HCC patients. Three patients demonstrated a partial response and six showed a stable disease for more than 6 months, and only one discontinuation for toxicity (51).
Thalidomide was originally developed as a treatment for insomnia and is highly effective in myeloma. Its precise mechanism of action is unknown. It appears to have multiple actions: inhibition of angiogenesis (anti-FGFR, VEGFR), inhibition of growth and survival of stromal cells, alteration of production/activity of cytokines [cyclooxygenase 2, tumour necrosis factor α (TNFα) – role in cancer-induced cachexia – interleukins 4, 5, 6, 8, 10 and 12], alteration of the expression of adhesion molecules and stimulation of T cell. Thalidomide has been given to HCC patients, but its efficacy appears to be weak, and annoying side effects such as fatigue and somnolence limit its administration in cancer patients (52).
Sunitinib is a molecule that inhibits multiple tyrosine kinase of receptors: VEGFR1-2, FLK, PDGFR, FLT3 and c-kit. The anti-angiogenic effect of sunitinib is mainly mediated through VEGFR and PDGFR protein kinase inhibition, and also via the direct antiproliferative effect through blockage of FLT3 and c-kit receptors. Sunitinib is already approved for the treatment of gastrointestinal stromal tumour and renal cell carcinoma. The results of the available phase II trials indicate a more intense antitumoral effect than sorafenib when assessing the appearance of tumour necrosis rather than the tumour size. A 50 mg/day dosage is associated with major side effects such as bleeding and death (53), and a lower dose seems to reduce this incidence but does not really eliminate it (54). The positive data with sorafenib have prevented a comparison vs. placebo and major, strong data regarding safety and efficacy as measured by TTP and survival in phase 2 should be available before running a head-to-head comparison can be made. Indeed, if sorafenib is the standard of care, with the available data, the expected trial would take place in second line.
NGR-hTNF is a novel agent that selectively binds to aminopeptidase N/CD13, highly expressed on tumour blood vessels. A phase II trial enrolled 16 patients with progressive HCC. NGR-hTNF was given at 0.8 mcg/m2 for 1 h via an intravenous infusion every 3 weeks. Thirty-six per cent of patients showed disease stabilization for more than 4 months. The median PFS was 2.4 months, and no grade 3–4-related toxicities were noted (55).
Drugs targeting other pathways and other putative hepatocellular carcinoma treatments
Microtubule inhibitors, patupilone and vinflunine, have yielded poor results in phase II studies (56). Bortezomib, an inhibitor of protein turnover, has also yielded disappointing data in a phase II study in association with doxorubicin (57). A putative efficacy of ADI-PEG20 demonstrated in a phase I study needs to be explored in a phase II study.
Belinostat (PXD101), an inhibitor of histone deacetylase, was given to 12 HCC patients. The phase I trial indicated a very good tolerance, with a MTD not being reached at a dose level of 1200 mg/m2/day; a phase II trial is ongoing.