Sorafenib in Hepatocellular Carcinoma

: Hepatocellular carcinoma (HCC) is the ﬁ fth most common malignancy worldwide with approximately 500,000 new cases per year, ranking as the third cause of cancer deaths. Until the very recent past no systemic therapy was available for patients with advanced HCC. However, advances achieved in the understanding of the molecular cancer biology and the subsequent introduction of molecularly targeted agents have disclosed promising and interesting possibilities for the treatment of this highly vascularized tumor in which the inhibition of angiogenesis is likely to represent the main potential therapeutic target. In this regard, sorafenib, an oral multikinase inhibitor endowed with signi ﬁ cant activity against several tyrosine-kinase receptors involved in blood vessel development, has shown to be very active in the treatment of advanced HCC. This review makes a survey of pharmacology, mode of action, pharmacokinetics, when


Introduction
Hepatocellular carcinoma (HCC) represents one of the leading solid organ malignancies worldwide, accounting for 6% of all tumor types. Approximately 60% to 80% of incidence relates to its common association with cirrhosis from heavy and long-lasting alcohol consumption, chronic hepatitis virus infections (hepatitis B virus [HBV] in Asia and Africa and hepatitis C virus [HCV] in the Western countries), and afl atoxin exposure. Consequently, the management of this tumor becomes even more diffi cult because of the presence of this underlying liver disease. 1,2 Surgery is the mainstay of HCC treatment 3 but at presentation, more than 80% patients present with multicentric HCC and advanced liver disease or comorbidities that restrict the therapeutic measures to BSC. 4 As far as the treatment of advanced disease is concerned, to date no single agent or combination chemotherapy have demonstrated any advantage in terms of both overall survival and quality of life. Also the hepatic arterial infusion and chemoembolization, which in some instances resulted in an increase of response rate, did not show any benefi t in terms of overall survival. 5 Lastly, no treatment has been identifi ed for severe advanced disease whose prognosis is very poor and is measured in months. 2 In the very recent past, however, the signifi cant advances achieved in the understanding of cell-cycle regulation and tumorigenesis have identifi ed new targets for therapeutic intervention and have disclosed new and very promising perspectives for the treatment of this disease. In particular, the potential targets for HCC include the vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), and hepatic/c-met signalling pathways. Neovascularization is one of the most important factors and there is a close relationship between VEGF expression and HCC vascularization compared to non-cancerous liver tissue. 6 It has been demonstrated that VEGF is overexpressed in HCC and correlates with a decrease in disease-free survival and overall survival, and with an increase of portal-vein invasion which are associated with later stage of disease, high grade and poor prognosis. Moreover, it has been observed that also Ras mutations are overexpressed in HCC tumors [6][7][8] demonstrated that dual inhibition of Raf and VEGFR2 reduces cells growth and vascularization of hepatocellular carcinoma in experimental model. 9 Sorafenib (Nexavar®, Bayer/Onyx), an oral mutikinase dual-action inhibitor able to target Raf kinase and vascular endothelial growth factor receptors so preventing both tumor cell proliferation and angiogenesis, has been extensively studied in HCC. Following a series of preclinical and clinical studies carried out in this tumor type which clearly demonstrated a high level of activity, the unquestionable results of two pivotal randomized phase III clinical trials have acknowledged sorafenib as the only and the unique drug presently available and recommended for the treatment of advanced HCC. 5,10 This review summarizes the main characteristics of sorafenib running through the series of results achieved in the management of HCC.

Pharmacology, Mode of Action and Pharmacokinetics of Sorafenib
Due to the lack of understanding of the molecular mechanisms underlying tumorigenesis, cytotoxic chemotherapy has represented for many years the only medical approach to cancer treatment. 7 In the last decades of the past century, however, advances in molecular biology have led to the identifi cation and characterization of genetic elements (gene mutation, gene amplifi cation and/or overexpression) whose normal function is to control cell growth and/or differentiation, which, when altered, play a fundamental role in human cancer pathogenesis. This occurs through the activation of abnormal complex signaling networks mainly involving growth factors receptors and signal-transduction pathways. Consequently, large drug discovery programs have been activated and devoted to the identifi cation of biologically specifi c therapies able to target the cancer-specifi c pathways, 1 also in the assumption that the new targeted agents, theoretically sparing normal cells, could provide a higher therapeutic index on account of better safety benefi ts over the standard cytotoxic chemotherapy. 7 The fi rst proof of concept that gene is a valid anticancer target goes back to 1989 when it was demonstrated that the disruption of raf1 gene inhibits the growth of a large number of human tumors: as a consequence of this, the screening of drugs with Raf1 kinase inhibitory activity was initiated and developed.
Early investigations carried out in cell lines have shown that sorafenib directly inhibits downstream Raf kinase isoforms (wild-type Raf1, B-Raf and mutant b-raf V600E). The subsequent discovery that the drug is also able to block the autophosphorylation of several tyrosine-kinase receptors such as VEGFR1, 2 and 3, PDGFR, c-Kit and RET, has led to the settlement that sorafenib is a dual-action drug capable to inhibit both tumor-cell proliferation and tumor angiogenesis. 7 It is likely that the mechanism of action of this drug has not yet been completely elucidated since, as it has been demonstrated in a wide range of tumor models, it seems able to increase also the rate of apoptosis. 11,12 • Sorafenib demonstrated high interpatient PK variability, which is not explained by age, race, gender, or body weight 1 • Safety data at 400 mg bid were acceptable despite the interpatient PK variability 1 Starting from the very early stages of the disease, HCC appears as a complex heterogeneous hypervascular tumor in which the overexpression of proangiogenic factors like VEGF, PDGF and angiopoietin-2 has been largely demonstrated. Experimental studies have shown that targeting angiogenesis is a top priority objective in this tumor in which vascular endothelial growth factor (VEGF) is responsible for the enhancement of development and of metastases. 13,14 Besides that, there is evidence of aberrant activation of several signaling cascades such as EGFR, Ras/extracellular signal-regulated kinase, phosphoinositol 3-kinase/ mTOR, hepatocyte growth factor/mesenchymalepithelial transition factor, Wnt, Hedgehog, and apoptotic signaling. 8 Additionally, the HCV core proteins, frequently involved in cellular transformation, elicit high basal Raf-1 activity in hepatocytes thus increasing the risk of neoplastic conversion. 15,16 Therefore, on account of its ability to inhibit angiogenesis and RAF kinase pathway, sorafenib has been identifi ed as a suitable and ideal candidate for studies in HCC.
The pharmacokinetic evaluation of sorafenib in humans was carried out in 60 patients with advanced refractory solid tumors by testing single and repeated orally administered doses of the drug ranging from 100 to 800 mg bid. Notwithstanding an inter-individual variability of the pharmacokinetic profi le likely due to different individual levels of absorption of the drug and/or of entero-hepatic circulation, results showed that there was no dosedependent relationship in pharmacokinetic parameters when testing the doses of 400, 600 and 800 mg considered as the therapeutic ones. Maximum mean values of AUC and C max were reached between 2.5 and 12.5 hours following the administration of a single dose of 600 mg bid, with only marginal differences as compared to the dose of 400 mg bid. 17 Further increase of the dose to 800 mg bid did not result in any increase of the AUC 0-12 and C max values. Substantial accumulation of sorafenib in plasma (2.5 to 7-fold) has been observed after repeated administrations: steady-state plasma concentrations were achieved after seven days. [17][18][19] Food intake before dosing slightly modifi es the pharmacokinetic parameters: on the contrary, the fat content of the meal seems to have some infl uence. 18,19 Mean AUC and C max values are reduced of about 29% and 38% respectively when sorafenib is administered following a high-fat meal, while an increase of about 14% of the AUC and a decrease of 17% of C max have been observed when sorafenib is given with a moderate-fat meal. 19 Sorafenib highly binds to plasma proteins (99.5%) with a binding affi nity which is linear and independent of the drug concentration. 19 It primarily binds to serum albumin and, to a lesser extent, to α-globulins, β-globulins and LDL, while does not bind to γ-globulins and α1-acetic glycoprotein. 20 Metabolism of sorafenib has been extensively evaluated both in in vitro and in vivo studies. Evidences coming from in vitro investigations carried out on human liver cell microsomes indicate that the fi rst oxidative metabolism is primarily mediated by CYP3A4. Moreover, evaluations in human hepatocytes have also specified that sorafenib is metabolized through a glucuronidation pathway involving UGT1A9. Sorafenib is a moderate inhibitor of CYP2C19, CYP2D6, CYP3A4, CYP2B6, CYP2C8 and CYP2C9, and the evaluation of the in vitro enzyme induction potential of sorafenib showed no activation of the CYP2C19 and CYP23A enzyme activity.
Studies in vivo using radiolabelled sorafenib showed that the drug is predominantly excreted with feaces (approximately 77%) and in a lower proportion by renal route (19%). No data concerning the pharmacokinetics in children are presently available. As regards to adults, no dose adjustments are needed according to age, gender or renal impairment. 21 On the contrary, ethnicity has shown to infl uence the pharmacokinetic parameters: a comparison between Japanese patients and Caucasian patients indicates that in the former cohort AUC values are 45% lower with a signifi cant overlapping in the range of exposures observed in the two groups. 21,22 Effi cacy Studies Preclinical studies Notwithstanding sorafenib was initially recognized as a CRAF inhibitor, 23 additional studies have also demonstrated that it is a potent inhibitor of other kinases such as BRAF (wild-type and mutant), VEGFR2, VEGFR3, PDGFR-β, FLT3, Ret and c-Kit. 23,24 This evidence clearly appeared in preclinical models where a dose-dependent activity against a wide range of tumor types, including HCC, in terms of cell-growth inhibitory effects, induction of apoptosis, and downregulation of the antiapoptotic protein Mcl-1 through a Raf/MEK/ERK-independent mechanism was observed. 11,24 As far as HCC is concerned, preclinical studies have demonstrated that Raf/MAPK-ERK kinase (MEK)/extracellular signal regulated kinase (ERK) pathway plays an important role. The over-expression of activated MEK1 in HCC cell lines, which enhances tumor growth and survival by preventing apoptosis, suggested that targeting the MEK signalling pathway could be an appropriate approach to the treatment of HCC. 25,26 An evaluation of the direct effects of sorafenib carried out in vitro n.     in two different HCC cell lines showed that sorafenib is endowed with antiproliferative activity and induces cell apoptosis in a dose-dependent way. In addition, sorafenib also inhibits the MEK and ERK phosphorylation. Subsequent evaluations of the effects of sorafenib carried out in vivo in mice using a xenograft model showed that the drug induces a signifi cant and dose-dependent tumor growth inhibition achieving partial tumor regression in about 50% of animals. 11 The entirety of these evidences provided the rationale for clinical studies with sorafenib in human HCC.

Clinical studies Preliminary remarks
Since in most cases of HCC the concomitant presence of cirrhosis substantially affects the evaluation of the stage of the disease and, afterwards, the choice and the outcome of treatment, the measure of the extent and severity of hepatic dysfunction represents a crucial aspect which cannot be left out. Consequently, the TNM cancer staging system cannot be considered adequate and satisfactory to properly defi ne a situation resulting from the concurrence of quite a lot of factors. Therefore, several systems for a better defi nition of the stage of the disease and prognosis have been proposed. Even though in a retrospective analysis of patients with advanced HCC carried out by the Memorial Sloan-Kettering Cancer Center the CLIP (Cancer of the Liver Italian Program) and GRETCH (GRoupe d'Etude et de Traitement du Carcinoma Hépatocellulaire) scoring systems seemed to satisfactorily fulfi ll the requirements for a valid defi nition, 2 nowadays the Barcelona-Clinic Liver Cancer (BCLC) classifi cation, which links tumor stage with treatment strategy, has been acknowledged and endorsed by AASLD (American Association for the Study of Liver Diseases) and EASL (European Association for the Study of the Liver) as the standard one for clinical management of HCC. 3 Another very important aspect, which deserves to be clarifi ed and which calls for the choice of new more suitable and reliable alternatives so far not yet undeniably identifi ed and still matter of evaluation and discussion, relates to the assessment of tumor response when using the molecularly targeted agents. Likewise other targeted agents, in all studies carried out with sorafenib in different tumor types RECIST (Response Evaluation Criteria in Solid Tumors) criteria, which grade the tumor response on the basis of the extent of tumor shrinkage, have been used: this resulted in a rather limited rate of complete and/or partial responses and in a striking rate of stable diseases. This goes off because the RECIST criteria, originally developed to assess responses to cytotoxic drugs, are not appropriate as indicators of activity of sorafenib and of other targeted agents whose mechanism of action lies in a block of cell growth without necessarily inducing cell lysis. As a matter of fact, tumor scans have shown that most cases graded as stable disease on the basis of the lack of tumor shrinking represent on the contrary a consistent response to treatment as a necrotic area inside the tumor burden, usually spreading from centre to the periphery, has developed.
Phase I Two single-agent phase I trials 27 evaluated sorafenib at doses ranging from 50 to 800 mg/day in 69 and 27 patients with a variety of advanced refractory solid tumors. Among 6 evaluable patients with HCC, 2 stable disease lasting Ͼ6 months, 1 stable disease lasting Ͼ1 year and 1 partial response Ͼ6 months have been observed. Even if the majority of patients experienced at least one adverse event, the grades of toxicities were mostly mild to moderate. 18 Another phase I single-agent disease-oriented trial was carried out in 27 Japanese patients with advanced unresectable HCC (Child-Pugh class A or B). They received at first a single dose of sorafenib of 200 mg (n = 13) or 400 mg (n = 14) followed by a 7-day rest period: subsequently, the same doses of sorafenib (200 or 400 mg bid) have been administered for 28 consecutive days in 4-week cycles. Out of 27 patients evaluable for effi cacy, of whom 10 treated with 200 mg bid and 11 with 400 mg bid, 1 (4%) achieved a partial remission (PR) and 21 (77.8%) had stable disease (SD) evaluated according to RECIST criteria. Median Time To Progression (TTP) was 4.9 months. Median Overall Survival (OS) in all intent-to-treat 27 patients was 15.6 months. The authors concluded that, in Japanese patients with advanced HCC, sorafenib at the doses of 200 mg and 400 mg bid is generally well tolerated showing a consistent anti-tumor activity. 22 An open-label, uncontrolled, phase I trial evaluating the combination of sorafenib with doxorubicin also showed a good level of effi cacy of this regimen in the treatment of patients with advanced HCC. 28 Phase II An open-label, multicentre, phase II trial evaluating sorafenib single agent at the dose of 400 mg bid continuous dosing was carried out in 137 previously untreated patients with advanced, inoperable HCC Child-Pugh class A or B. The evaluation of patients on the basis of an independent assessment showed PR in 3 (2.2%) patients, minor responses in 8 (5.8%) and SD lasting for at least 16 weeks in 46 (33.6%). Median TTP was 5.5 months and the OS was 9.2 months. 1 The latter fi gure appears particularly interesting since it compares favorably with those of other studies evaluating combination regimens with conventional cytotoxic agents (PIAF [cisplatin, interferon alpha-2b, doxorubicin, fl uorouracil] or doxorubicin plus cisplatin) in HCC patients 2,3 where median overall survival rates accounting for 8.9 and 7.3 months respectively have been reported. 1 In addition, in a small group of patients (33), pretreatment tumor phosphorylated ERK (pERK) levels have been correlated with the TTP: results show that patients who had tumors expressing more intense pERK staining had also a longer TTP. 1 Another phase II, randomized, double-blind, placebo-controlled study evaluated the combination doxorubicin + sorafenib vs doxorubicin + placebo in 96 previously untreated patients with advanced HCC Child-Pugh class A or B. Patients were randomly allocated to receive either doxorubicin (60 mg/m 2 every 3 weeks up to six cycles) plus sorafenib 400 mg bid (n = 47), or doxorubicin plus placebo (n = 49). After six cycles of doxorubicin and in the absence of disease progression patients could decide to continue the treatment with either sorafenib or placebo alone until disease progression, withdrawal or death. Median TTP, the primary endpoint, was 8.6 months for patients in the doxorubicin + sorafenib arm vs. 4.8 months for those in the doxorubicin + placebo arm. Both median Progression Free Survival (PFS) (6.9 vs. 2.8 months, respectively; HR 0.57; p = 0.012) and median OS (13.7 vs. 6.5 months, respectively; HR 0.45; p = 0.0049) were signifi cantly prolonged in the doxorubicin + sorafenib group as compared to the doxorubicin + placebo group. The Overall Response Rate (ORR) was similar in both groups (4% vs. 2%, respectively), SD rate achieved in the doxorubicin + sorafenib group was 77% vs. 55% in the doxorubicin + placebo group. The toxicity profi les were similar in the two arms of the study and included toxicities commonly seen with singleagent doxorubicin and sorafenib. Grade 3-4 toxicities included fatigue (15%) and neutropenia (50%) in both arms. Sorafenib-related toxicity included grade 3-4 diarrhea (11%) and grade 3-4 hand-foot syndrome (9%) in the combination arm. More left ventricular dysfunctions in the doxorubicin plus sorafenib arm have been observed, which accounted for 19% for all grades and 2% for grades 3-4. 2,29 Phase III Two pivotal large phase III trials have been carried out to defi nitely assess the effi cacy and safety of sorafenib in advanced HCC.
The study SHARP (Sorafenib HCC Assessment Randomized Protocol) was an international, multi centre, randomized, double-blind, placebocontrolled trial comparing sorafenib 400 mg bid continuous dosing vs. placebo in patients with ECOG Performance Status 0-2 and Child-Pugh class A not previously treated with systemic therapy. 5 Six hundred-two patients have been randomized and enrolled in 121 sites scattered throughout 21 countries worldwide (Americas, Europe, and Australia/ New Zealand). The primary objectives of the study were OS and time to symptomatic progression (TTSP), the latter assessed using the Functional Assessment in Cancer Hepatobiliary Symptom Index Time to Symptom Progression (FSHI-8-TSP) scoring system. TTP, disease control rate and safety were secondary endpoints. Before randomization, patients have been stratifi ed according to geographical region, ECOG Performance Status (0 vs. 1-2) and tumor burden (presence vs. absence of macroscopic vascular invasion and/or extrahepatic spread). Overall, 299 patients have been randomized to receive sorafenib and 303 to receive placebo. Approximately 50% of the enrolled patients had either hepatitis C virus or hepatitis B virus as underlying etiology, and 26% had alcohol related cirrhosis. Patients with underlying Child-Pugh class A cirrhosis accounted for 95% and 98% in the sorafenib and placebo groups, respectively. More than 80% of the patients in the sorafenib arm had portal vein invasion and were classifi ed as BCLC stage C; 20% of them had extra-hepatic metastases. Most patients had mild to moderate performance status. Based on a planned interim analysis carried out in February 2007 when 321 forecasted events occurred, the Data Monitoring Committee (DMC) concluded that the trial met the primary endpoint since the OS in the group of patients treated with sorafenib turned out statistically signifi cant superior as compared to the one of patients receiving placebo (10.7 months for sorafenib vs. 7.9 for placebo; HR = 0.69; p = 0.0005), with a 31% reduction of risk of death. The TTP achieved in patients treated with sorafenib was almost double as compared to the group of patients treated with placebo (5.5 months vs. 2.8; HR = 0.58; p Ͻ 0.001) and also the disease control rate (Complete Responses + Partial Responses + Stable Diseases Ն6 weeks) was signifi cantly higher in patients treated with sorafenib (43 vs. 32%; p = 0.002). As far as TTSP is concerned, the study failed to demonstrate any statistically significant difference between sorafenib and placebo (median 4.1 vs. 4.9 months, respectively; HR = 1.08; p = 0.77); the most likely reason for this lack of difference, could be ascribable to the reporting of drug-related toxicities which may have infl uenced the assessments of the FHSI-8 scoring system, mainly because very often it prevented from making a clear distinction between drug-related toxicities and tumor symptom-related events. 5 The DMC also underlined the lack of imbalance in the serious adverse events rate in the two experimental groups. On the light of these results, the DMC stated that all patients in the placebo group could have access to sorafenib treatment and, on the basis of these conclusions, the trial was stopped earlier.
The ASIA-PACIFIC study is an international phase III, multicentre, randomized, double-blind, placebo-controlled trial which has evaluated sorafenib vs. placebo in 226 previously untreated patients with advanced HCC or primary liver cancer coming from sites across China, Taiwan and Korea.
Differently from the SHARP trial, the main etiological factors in this study were Asian ethnicity and hepatitis B virus. Following stratification according to macroscopic vascular invasion and/or extra-hepatic spread, ECOG PS and geographic region, patients have been randomized in a 2:1 ratio either to sorafenib 400 mg bid (n = 150) or to placebo (n = 76) treatment. Study endpoints included OS, TTP, PFS, TTSP, disease control rate (CR + PR + SD Ն 6 weeks) and safety. The primary endpoint was not defi ned. Similarly to SHARP study, also the ASIA PACIFIC trial showed a superiority of sorafenib as compared to placebo in terms of OS, PFS, TTP and disease control rate, further demonstrating the effi cacy of this drug also in patients with different ethnicity and different leading etiological factors. Results, reviewed and validated by an independent DMC, indicated statistically signifi cant differences as regard TTP (2.8 months for sorafenib vs. 1.4 for placebo; HR 0.57; p Ͻ 0.001), OS (6.5 months vs. 4.2; HR 0.68, p = 0.014) and disease control rate (35% [95% CI [28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43][44] vs. 16% [95% ). Also in this study the TTSP evaluation of the two experimental groups did not show any statistically signifi cant difference (3.5 months for sorafenib vs. 3.4 for placebo; HR = 0.90; p = 0.498). 10 The observation that the absolute benefi t in the ASIA PACIFIC study was smaller as compared to SHARP trial could be explained by the fact that patients in the oriental trial showed a poorer performance status and more advanced tumor stage as exemplifi ed by a higher rate in extra-hepatic spread.

Safety
Similar rates of drug-related adverse events in terms of both incidence and category, the most common being represented by dermatological, constitutional and gastrointestinal symptoms, have been observed in the phase II and phase III studies carried out in patients with advanced HCC treated with sorafenib 1,5,10 In the large openlabel phase II study where 137 patients have been treated, no grade 4 drug-related adverse events have been observed. As for the grade 3 events recorded in Ն10% of all patients, they accounted for 5.1% hand-foot skin reactions (HFSR), 0.7% rash/desquamation, 9.5% fatigue, 8.0% diarrhea, 1.5% anorexia and 0.7% stomatitis.
In the phase III SHARP trial, the overall incidence of adverse events of any grade and from any cause was 80% in the sorafenib group and 52% in the placebo group; they mainly consisted of gastrointestinal, constitutional and dermatologic symptoms. On the contrary, the overall incidence of serious adverse events from any cause was similar in the two study groups accounting for 52% in the sorafenib group and 54% in the placebo group: in relation with the particular features of the disease they were mainly represented by hepato-biliary events (11% and 9%, respectively), hemorrhagic events (9% and 13%), variceal bleeding (2% and 4%), renal failure (Ͻ1% and 3%). The most frequent drug-related grade 3 adverse events included diarrhea (8% in the sorafenib group vs. 2% in the placebo group, p Ͻ 0.001), HFSR (8% vs. Ͻ1%, p Ͻ 0.001), abdominal pain (2% vs. 1%, p = 0.17), hypertension (2% vs. Ͻ1%, p = 0.28) and cardiac ischemia or infarction (3% and 1%). No grade 4 drug-related adverse events among any of these categories have been reported. Grade 3-4 laboratory abnormalities occurred with similar frequencies in the two study groups, with the exception of grade 3 hypophosphatemia (11% in the sorafenib group vs. 2% in the placebo group, p Ͻ 0.001) and grade 3 or 4 thrombocytopenia (4% in the sorafenib group vs. Ͻ1% in the placebo group, p = 0.006).
The drug-related adverse events recorded in the ASIA-PACIFIC study are in line in terms of incidence and category with those observed in the SHARP study, being the grade 3-4 represented by HFRS (11%), diarrhea (6%), fatigue (9%), and hypertension (2%).
Worth nothing is the fact that, with exception of the hepatobiliary adverse events strictly related to the presence of liver disease, all the drug-related events appeared in the studies on HCC correspond in terms of both incidence and type to those gathered when using sorafenib in other studies in different tumor types such as the treatment of renal cell carcinoma e.g. phase III TARGET study. 30 It is likely that some adverse events typical of sorafenib and of other targeted agents could be related to the particular mechanism of action of these drugs. Dermatological toxicity, HFRS in particular, represents one of the major features following the treatment with sorafenib: the mechanism of onset has not yet fully understood but almost certainly we are dealing with a side effects strictly connected with the anti-kinase activity. 2,3 Analysis of cumulative event rate from the phase III pivotal studies carried out with sorafenib in all tumor types suggests that rash and HFSR usually appear during the fi rst 6 weeks of treatment with sorafenib and they get worse as the treatment goes on. 30 Management of these dermatologic toxicities may include topical therapies for symptomatic relief, temporary treatment interruption and/or drug dose modifi cations, or, in case of increase of severity or persistence, permanent discontinuation of sorafenib. [31][32][33] In rare instances, yellow skin discoloration, mainly on hands and feet, has also been observed during sorafenib treatment: however, generalized skin discoloration has not been reported. 34 Also the increase of incidence of arterial hypertension seems to be related to the mechanism of action of tirosyne-kinase inhibitors: hypertension has been reported in a fairly considerable percentage of patients treated with sorafenib (5% of patients treated in the SHARP study and 12% in the TARGET study). It occurred early in the course of treatment, it was usually mild-to-moderate, and, in the majority of cases, it was controlled with standard anti-hypertensive therapies. 30,35 Also cardiac toxicity seems peculiar of drugs targeting tyrosine-kinase patways. In the SHARP study, ischemia/infarction was observed in 3% of patients treated with sorafenib and in 1% in the placebo group: similarly, in the TARGET study, the incidence of treatment-emergent cardiac ischemia/infarction events was higher in the sorafenib group (2.9%) as compared to placebo group (0.4%). As regards other signs of cardiac toxicity, the incidence of decreases of LVEF values following treatment with sorafenib has been very little: only in a single patient a drop of 10% of LVEF was reported. In the sorafenib safety database, the incidence of congestive heart failures is Ͻ2%. 36

Quality of Life
No data on quality of life of HCC patients treated with sorafenib are presently available, mainly on account of the fact that a reliable evaluation of this parameter in patients bearing this tumor type is really extremely diffi cult to carry out. As previously mentioned, in both phase III pivotal studies in HCC 5,10 the assessment of the TTSP using the FSHI-8-TSP scoring system failed to demonstrate any difference between the cohorts of patients treated with sorafenib and placebo. Therefore, at least for the time being, the only evidences it is possible to rely on are the level of tolerability of the treatment and its level of acceptance by patients. In this regard, data of the SHARP trial, which show that the rate of discontinuation of the study drug due to adverse events was similar in the two study groups (38% vs. 37%), could be taken as a tool to indicate that sorafenib is a well tolerated drug scarcely affecting the patients' quality of life. 5

Conclusions
To date, sorafenib represents the fi rst and the only systemic therapy able to prolong survival in patients with advanced HCC: evidences coming from pivotal clinical trials led to the approval of this drug by the International Regulatory Authorities (FDA and EMEA) as the new reference standard treatment for patients with advanced HCC. 3 The clinical benefi t obtained with sorafenib in HCC with a decrease of risk of death of 31% is similar in size to the ones obtained using other molecular targeted therapies in other tumor types such as trastuzumab in breast cancer, bevacizumab in colorectal cancer, erlotinib in non-small-cell lung cancer, which induced decreases in the hazard of death ranging from 25% to 35%. 8 Current knowledge leaves unsolved some important issues.
It is necessary to increase the Overall Survival benefi t highlighted in the phase III trials, though an 0.69 HR it is comparable with the above mentioned therapies a 3 months increase in survival is not changing this disease natural history. Combination trials with sorafenib and other target agents have been planned in advanced HCC. 3 It is also important to plan sorafenib development in adjuvant therapy after potentially curative treatments, as resection and local ablation, due to the crucial need of an adjuvant therapy for high recurrence rate in this tumor. 3 Another important point, is the lack of prospective trials on biomarkers for sorafenib . Two retrospective analysis from HOPE 1 and SHARP 5 exclude VEGFR as possible biomarker; so it is diffi cult to understand if a positive correlation between sorafenib clinical effect and HCC vascularization exists. A biomarker should anticipate the real effi cacy of a drug on a certain patient. This should optimize the public health expenditure, given the high cost of new biological therapies. Searching a good biomarker candidate it is diffi cult due to the lack of knowledge of sorafenib mechanism of action in human HCC.
No doubt however that the positive results of sorafenib represent an advancement in the care of patients for whom no effective therapy was previously available: nonetheless, this represents only the fi rst step for a further development of this drug in advanced disease in combination with chemo-embolization or other molecularly targeted therapies, and, almost certainly, in the near future also in the adjuvant setting in earlier stages of disease after surgery. 8