Phase 2 open-label study of single-agent sorafenib in treating advanced hepatocellular carcinoma in a hepatitis B–endemic Asian population

Presence of lung metastasis predicts poor response

Authors


Abstract

BACKGROUND:

The current study was a phase 2 open–label study to evaluate the efficacy and tolerability of single-agent sorafenib in the treatment of advanced HCC patients in a hepatitis B–endemic Asian population.

METHODS:

Patients with advanced hepatocellular carcinoma (HCC) received sorafenib at a dose of 400 mg twice daily in 4-week cycles. Tumor response was assessed every 3 cycles using Response Evaluation Criteria in Solid Tumors criteria.

RESULTS:

Fifty-one patients were enrolled in the study and were treated with sorafenib for at least 12 weeks. The median age was 56 years (range, 28-79 years). Approximately 90% had hepatitis B virus–related HCC. Thirty-six (71%) patients had underlying Child-Pugh A cirrhosis, 13 (26%) Child-Pugh B, and 2 (3%) Child-Pugh C. Four (8%) patients achieved partial responses, and 9 (18%) patients had stable disease for at least 12 weeks. The median overall survival was 5 months (range, 4-17 months). Patients without extrahepatic spread, particularly without lung metastasis (P<.01), are more likely to benefit from sorafenib treatment. The most common toxicities were diarrhea (67%), malaise (55%), and hand-foot-skin reaction (54%). The majority of patients had transient liver function derangement. Patients with and without underlying portal vein thrombosis had similar therapeutic benefits and likewise shared a similar treatment-related toxicity profile with sorafenib treatment.

CONCLUSIONS:

Single-agent sorafenib demonstrates good efficacy and acceptable tolerability in treating an advanced HCC patient population in a hepatitis B–endemic area. The presence of lung metastasis predicts poor response to sorafenib in advanced HCC patients. Cancer 2009. © 2009 American Cancer Society.

Hepatocellular carcinoma (HCC) is the fifth most common cancer worldwide.1 There is great geographic heterogeneity in the incidence of HCC across the world, with the majority of cases found in Asian countries because of the high prevalence of endemic hepatitis B virus (HBV) infection.1 Although the long-term survival of patients with early HCC amenable to surgical treatment has improved in recent years, the prognosis of patients with advanced HCC remains dismal. Therefore, effective systemic therapies are urgently needed to improve the survival of advanced HCC patients. HCC is an aggressive, largely chemoresistant cancer with poor prognosis. To our knowledge, there is no convincing evidence to date that systemic chemotherapy improves overall survival in advanced HCC patients.2 Single-agent doxorubicin has been shown to produce a response rate of approximately 10% to 15%, but with no proven survival benefit.3 Because of the disappointing results of systemic therapy for advanced HCC, researchers have shifted their interests toward novel molecular-targeted agents.

Many molecular pathways play a pivotal role in the development and growth of HCC.4 Promising results have been shown in targeting the antiangiogenic pathways and the Raf/mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) pathways. Sorafenib, an oral multikinase inhibitor, blocks tumor growth by its antiangiogenic properties and by targeting tumor cell proliferation pathways, particularly the Raf/MAPK/ERK pathway.5 In the pivotal phase 3 randomized trial of sorafenib treatment in patients with advanced HCC conducted in Western patients,6 patients treated with single-agent sorafenib were shown to have a 44% increase in overall survival (OS) compared with placebo (hazards ratio, 0.69; P = .00058). However, in that study, only patients with Child-Pugh A cirrhosis were included, and very few (6%) of them had HBV-related HCC. Furthermore, most recruited patients had favorable clinical, biochemical, or radiologic parameters. Moreover, the preliminary results of the phase 3 study conducted by Cheng demonstrated an OS of 6.2 months in patients who received single-agent sorafenib for the treatment of advanced HCC in an Asian population.7 Nevertheless, similar to the design of the Sorafenib HCC Assessment Randomized Protocol Trial (SHARP) study,6 it only included Child-Pugh A patients with favorable clinical parameters. However, in daily clinical practice in most Asian countries, the patients with advanced HCC who are encountered are HBV carriers with suboptimal liver function, often Child-Pugh B or C cirrhosis. Even if they have Child-Pugh A cirrhosis, there are often some adverse tumor factors, such as underlying portal vein thrombosis (PVT). These patient populations are under-represented in the SHARP and Asian trial, and the efficacy and tolerability of single-agent sorafenib in these “real practice” patients are largely unknown.

We conducted a single-arm, phase 2 open-label study to evaluate the efficacy and tolerability of single-agent sorafenib in the treatment of advanced HCC patients presenting at our center between November 2006 and January 2008.

MATERIALS AND METHODS

Study Design

This was a phase 2, open-label, single-arm study conducted at Queen Mary Hospital, The University of Hong Kong. It was an investigator-initiated study and approved by the local ethic committee. Written consent was obtained from the patients before enrollment.

Patients' Eligibility

The diagnosis of HCC was based on increased serum α-fetoprotein level >400 ng/mL and typical imaging appearance according to the criteria of the European Association of Study of the Liver.8 Histology was only obtained by either fine–needle aspiration or biopsy in the case of diagnostic uncertainty.

Patients with advanced HCC who were not amenable to surgical resection, liver transplantation, or locoregional therapies were eligible to enter the current study. However, they had to have measurable lesions by computed tomography (CT) scan.

Treatment Regimen and Disease Assessment

All patients received single-agent sorafenib at a dose of 400 mg twice daily in 4-week cycles for at least 12 weeks. Patients were observed regularly every 2 weeks when they were receiving sorafenib. During the follow-up period after they were discontinued from sorafenib, assessment was performed every 4 weeks. Disease assessment was performed by CT scan approximately every 3 cycles (ie, 12 weeks). Magnetic resonance imaging and positron emission tomography with choline-acetate as radioisotope were performed in the case of diagnostic uncertainty. Response was determined by independent radiologists and classified according to Response Evaluation Criteria in Solid Tumors criteria.9 In this study, response was classified as complete response (CR), partial response (PR), stable disease (SD), or progressive disease. Patients who had achieved CR, PR, or SD were defined as achieving clinical benefits. Only patients who achieved clinical benefits continued sorafenib for another 3 cycles until radiologic evidence of disease progression or unacceptable treatment toxicities.

Dose Modification

Dose modifications were required for drug-related toxicities. For grade 3 of 4 toxicities, sorafenib was withdrawn until the toxicities improved to grade 2 or lower (grading determined according to the National Cancer Institute Common Toxicity Criteria). Afterward, sorafenib was reintroduced at a dose of 200 mg twice daily for 5 days and escalated back to 400 mg twice daily if well tolerated. Otherwise, sorafenib was continued as 200 mg twice daily.

Endpoints

The primary endpoint of the study was OS and the secondary endpoints included progression-free survival (PFS), tumor response rate, and drug tolerability.

Statistical Analysis

Survival analysis was computed by the Kaplan-Meier method. PFS was calculated from the date of commencement of sorafenib to the date of documented progression or death. OS was calculated from the date of commencement of sorafenib to the date of death or last follow-up.

Chi-square or Mann-Whitney U tests were used to assess potential clinical factors that might predict clinical benefit with sorafenib treatment. Potential predictive factors explored included age, sex, hepatitis B status, Eastern Cooperative Group (ECOG) performance status, Child-Pugh group, presence of vascular invasion, presence of extrahepatic disease, presence of lung metastasis, and prior systemic treatments.

RESULTS

Demographics

Between November 2006 and January 2008, 51 patients were recruited in this study, and all these patients had received sorafenib for at least 12 weeks. Table 1 shows the demographic data of these patients. The median age was 56 years (range, 28-79 years), and 88% were men. The majority of them had good performance status (95% had an ECOG performance status of 0 or 1, 5% had an ECOG performance status of 2). Forty-six (90%) patients were chronic hepatitis B carriers. In contrast, only 3 (6%) patients were chronic hepatitis C carriers. Hepatitis C serology was not available in 3 patients. One patient was a carrier of both hepatitis B and C viral infection. Thirty-six (71%) patients had underlying Child-Pugh A cirrhosis, 13 (26%) patients had Child-Pugh B cirrhosis, and 2 (3%) had Child-Pugh C cirrhosis. All patients had advanced HCC at the time of enrollment, and nearly half of them had extrahepatic metastases. The lung was the most common site of distant metastasis and was present in 20 (39%) patients, whereas PVT was present in 22 patients. Twelve patients had undergone prior liver resection for HCC, and only 1 patient had undergone liver transplantation. Four (8%) patients had received prior systemic therapy for advanced HCC.

Table 1. Demographic Data of the 51 Evaluable Patients
  1. ECOG indicates Eastern Cooperative Oncology Group; Hepatitis Bs Ag, hepatitis B surface antigen; Anti-HCV Ab, antihepatitis C antibody; AJCC, American Joint Committee on Cancer; TACE, transarterial chemoembolization; RFA, radiofrequency ablation; HIFU, high-intensity focus ultrasound.

Age, y
 Median56
 Range28-79
Sex
 Male45 (88%)
 Female6 (12%)
ECOG performance status
 035 (69%)
 113 (26%)
 23 (5%)
Hepatitis serology
 Hepatitis Bs Ag positive46 (90%)
 Anti-HCV Ab positive3 (6%)
Child-Pugh status
 A36 (71%)
 B13 (26%)
 C2 (3%)
Alpha-fetal protein
 ≤40026 (51%)
 >40025 (49%)
Disease stage at the time of study entry
 AJCC staging 
  II3 (6%)
  IIIA19 (37%)
  IIIB0 (0%)
  IIIC5 (10%)
  IV24 (47%)
 Distant metastases24 (47%)
  Lung20
  Bone3
  Adrenal3
  Brain2
 Invasion of major vessels:24 (47%)
  Portal vein invasion22
  Hepatic vein invasion1
  Inferior vena cava invasion1
Prior treatment
 Surgical treatment 
  Liver resection12 (24%)
  Liver transplantation1 (2%)
  Laparotomy2 (4%)
 Local ablative procedures 
  TACE20 (39%)
  RFA6 (12%)
  HIFU1 (2%)
  Alcohol injection1 (2%)
 Systemic therapy4 (8%)
 Radiotherapy3 (6%)
 Others 
  SIRTEX-yttrium1 (2%)

Treatment Efficacy

The median duration of sorafenib treatment was 3 months (range, 2.5-16 months), and 17 (33%) patients had the dose reduced by half during the treatment cycles because of treatment-related toxicities. Table 2 shows the efficacy results of the study. The overall response rate was 8%. There were no CRs observed, 4 (8%) patients had achieved PR, and another 9 (18%) patients had SD for at least 3 months. Therefore, 13 (26%) patients derived clinical benefits from sorafenib treatment.

Table 2. Best Response Assessment According to RECIST Assessment
ResponseNo. of Patients (%)
  1. RECIST indicates Response Evaluation Criteria in Solid Tumors.

Complete response0 (0)
Partial response4 (8)
Stable disease9 (18)
Progressive disease38 (74)

Survival Analysis

The overall median PFS and OS were 3 months (range, 3-17 months) and 5 months (range, 3-17 months), respectively (Fig. 1). Of the 13 patients who had demonstrated clinical benefits with sorafenib treatment, the median duration of treatment was 6 months (range, 3-16 months). Their median PFS and OS were 6 months (range, 3-17 months) and 10 months (range, 4-17 months), respectively. There was no statistical significant difference in OS between the responder and patients who had SD (13 vs 8 months; P = .3). In contrast, with respect to the 38 patients who did not achieve clinical benefits with sorafenib, their median PFS and OS were 3 months and 5 months, respectively. Of note, there were significant differences in PFS (6 months vs 3 months; P = .001) and OS (10 months vs 5 months; P = .006) between patients who demonstrated clinical benefits with sorafenib and patients who did not (Fig. 2).

Figure 1.

Kaplan-Meier curve of overall survival in patients receiving sorafenib for the treatment of advanced hepatocellular carcinoma is shown.

Figure 2.

Kaplan-Meier curve of overall survival in patients receiving sorafenib for the treatment of advanced hepatocellular carcinoma is stratified according to whether they derived benefit from sorafenib treatment.

Factors Predictive of Clinical Benefits With Sorafenib

Table 3 lists the results of univariate analysis of potential clinical factors predictive of clinical benefits with sorafenib treatment. Age, sex, ECOG performance status, hepatitis B carrier status, Child-Pugh group, invasion of major vessels and no prior systemic treatment did not predict clinical benefits with sorafenib. The absence of extrahepatic metastasis (P = .045), in particular the absence of lung metastasis (P<.01), significantly predicted clinical benefits with sorafenib treatment.

Table 3. Univariate Analysis of Potential Clinical Predictive Factors for Achieving Clinical Benefits While Receiving Sorafenib Treatment
Predictive FactorClinical Benefits (n=13), No. of Patients (%)No Clinical Benefits (n=38), No. of Patients (%)P
  1. ECOG PS indicates Eastern Cooperative Oncology Group performance status.

Age, y
 Median5656 
 Range32-6528-79.41
Sex
 Male13 (100)32 (84) 
 Female0 (0)6 (16).13
ECOG PS
 010 (77)25 (66) 
 1-23 (23)13 (34).45
Hepatitis B status
 Carrier13 (100)32 (84) 
 Noncarrier0 (0)6 (16).13
Child-Pugh group
 A7 (54)29 (76) 
 B or C6 (46)9 (24).125
Invasion of major vessels
 Yes7 (54)17 (45) 
 No6 (46)21 (55).57
Presence of extrahepatic spread
 Yes3 (23)21 (55) 
 No10 (77)17 (45).045
Presence of lung metastasis
 Yes1 (8)19 (50) 
 No12 (92)19 (50)<.01
Prior systemic treatment
 Yes0 (0)4 (11) 
 No13 (100)34 (89).58

Treatment-related Toxicities

Tables 4 and 5 show the details of treatment-related nonhematologic and hematologic toxicities, respectively, in the patients. With regard to nonhematologic toxicities, diarrhea (67%) was the most commonly encountered toxicity, followed by malaise (55%) and hand-foot skin reaction (HFSR) (54%). Moreover, the commonest grade 3 or 4 nonhematologic toxicities were diarrhea and HFSR, which occurred in 20% and 16% of patients, respectively. Hemorrhagic complications occurred in 3 patients, including epistaxis in 1 patient and upper gastrointestinal bleeding in 2 patients. With respect to hematological toxicities, thrombocytopenia (73%) was the commonest toxicity and was also the most frequently encountered grade 3 of 4 toxicity (10%). Table 6 shows the details of derangement in biochemistry of the patients on sorafenib. Notably, all patients except 1 had transient liver function derangement. Aspartate aminotransferase (AST) elevation was the most frequently affected liver biochemistry parameter and also the most common grade 3 toxicity, followed by bilirubin elevation. Grade 4 AST elevation was observed in 4% of patients, and grade 4 bilirubin elevation was observed in 6% of patients. In all these patients with grade 4 liver toxicities, the liver function gradually improved after sorafenib was discontinued. None of the patients developed liver failure or hepatic encephalopathy in the current study. There were no treatment-related deaths reported.

Table 4. Treatment-related Nonhematologic Toxicities
ToxicityAny Grade (%)Grade 1 or 2 (%)Grade 3 (%)Grade 4 (%)
  1. HFSR indicates hand-foot skin reaction.

Diarrhea34 (67)24 (47)10 (20)0 (0)
Malaise28 (55)24 (47)4 (8)0 (0)
HFSR27 (54)19 (38)8 (16)0 (0)
Alopecia19 (37)17 (33)2 (4)0 (0)
Rash15 (29)12 (24)3 (5)0 (0)
Abdominal pain12 (24)11 (22)1 (2)0 (0)
Hypertension11 (22)7 (14)4 (8)0 (0)
Nausea10 (20)8 (16)2 (4)0 (0)
Mucositis6 (12)6 (12)0 (0)0 (0)
Hemorrhage2 (4)1 (1)0 (0)0 (0)
Table 5. Treatment-related Hematologic Toxicities
ToxicityAny Grade (%)Grade 1 or 2 (%)Grade 3 (%)Grade 4 (%)
Thrombocytopenia37 (73)32 (63)5 (10)0 (0)
Leukopenia31 (61)31 (61)0 (0)0 (0)
Anemia31 (61)26 (51)2 (4)3 (6)
Neutropenia16 (31)15 (29)1 (2)0 (0)
Table 6. Treatment-related Derangement in Biochemistry
ToxicityAny Grade (%)Grade 1 or 2 (%)Grade 3 (%)Grade 4 (%)
  1. AST indicates aspartate aminotransferase; ALT, alanine aminotransferase; ALP, alkaline phosphatase.

AST50 (98)27 (53)21 (41)2 (4)
ALT40 (79)36 (71)4 (8)0 (0)
ALP39 (76)37 (73)2 (4)0 (0)
Bilirubin36 (71)20 (39)13 (26)3 (6)
Creatinine5 (10)4 (8)1 (2)0 (0)

Outcomes of Patients With PVT

Among 22 patients with PVT, there were 20 male and only 2 female patients. The median age was 53 years (range, 32-68 years). Seven patients had clinical benefits with sorafenib; among them, 4 patients had PR, and the other 3 patients demonstrated SD. Thus, 32% of patients with PVT achieved clinical benefits with sorafenib treatment. There were no significant differences between patients with and without PVT in terms of clinical benefits (32% vs 21%; P = .37) and OS (5 months vs 6 months; P = .65). Moreover, there were no differences in the incidence of grade 3 of 4 nonhematologic (36% vs 52%; P = .28) and hematologic toxicities (14% vs 28%; P = .23) between patients with and without PVT, respectively.

Analysis of Child-Pugh B and C Cirrhosis Patients

Fifteen patients with Child-Pugh B or C cirrhosis were included in the analysis. There were no significant differences between Child-Pugh A and Child-Pugh B/C patients treated with sorafenib in terms of clinical benefits (19% vs 40%; P = .125) and OS (5.5 months vs 5 months; P = .24). In addition, there were no significant differences between Child-Pugh A and B/C patients with regard to grade 3 or 4 hematologic toxicities (17% vs 33%; P = .18) and grade 3 or 4 nonhematologic toxicities (47% vs 47%; P = .97) after sorafenib treatment. More importantly, there was no significant difference in the incidence of grade 3 or 4 liver function derangement between patients with Child-Pugh A and Child-Pugh B/C cirrhotic patients treated with sorafenib (56% vs 73%, P = .236).

DISCUSSION

In the current study, single-agent sorafenib demonstrated fairly good efficacy and acceptable tolerability in treating advanced HCC patients in a hepatitis B–endemic area. The 8% response rate and 18% disease-stabilization rate achieved in this study is encouraging and is comparable to the results reported in the phase 2 study conducted by Abou-Alfa et al10 and the phase 3 randomized controlled trial (SHARP study).6 Although the disease-stabilization rate of 35% is higher in the Asian study, this rate refers to disease stabilization for more than 28 days, which is in sharp contrast to the disease-stabilization rate of at least 3 months in the current study. Conversely, both PFS (3 months) and OS (5 months) of patients in the current study are apparently inferior to the survival reported in the aforementioned phase 2 study conducted by Abou-Alfa et al10 and the SHARP study.6 This is expected, as the study population is different, and most patients enrolled in our study had poor overall prognosis because of advanced tumors (such as PVT) or Child-Pugh B/C cirrhosis. Such patients were mostly excluded from the SHARP and Asian studies.

In the SHARP study, 602 patients with biopsy-proven advanced HCC were recruited, and the median OS of patients on sorafenib and placebo was 10.7 months and 7.9 months, respectively. Conversely, in the Asian study, a total of 226 patients were recruited and randomized in a 2:1 fashion. The median OS of patients receiving sorafenib and placebo was 6.2 months and 4.1 months, respectively. In fact, the outcomes of patients in the Asian study were worse than patients in the SHARP study because of the enrollment of more patients with worse prognosis. Notably, the study population in the current trial has even poorer prognosis than the Asian study because the median OS of this patient population in our historical cohort is only 2.3 to 2.6 months.11 Impressively, the PFS of 3 months and OS of 5 months achieved in the current study are still comparable with the findings of the Asian study and are far better than our historical results.

More importantly, in patients who have achieved clinical benefits with sorafenib, there was a 4-fold increase in overall survival when compared with our historical cohort. Interestingly, even in patients who did not demonstrate any clinical benefits with sorafenib, OS was substantially better compared with our historical cohort results. It may suggest that even those patients who had no demonstrable clinical benefits with sorafenib might have some survival benefit from the treatment. This is an interesting phenomenon that has been observed recently in using targeting agents in the treatment of other solid tumors.12 With respect to patients with PVT, they did derive similar benefits from sorafenib when compared with patients without PVT. This finding further supports the finding from the SHARP study9 that advanced HCC patients with macrovascular invasion also benefit from sorafenib treatment. Notably, the 5–month OS achieved in patients with PVT in the current study is impressive, because in the literature this patient population usually has very short survival and grave prognosis.13

With regard to potential clinical or biological factors for predicting the possibility of clinical benefits of sorafenib treatment, it has been shown that pretreatment tumor phosphorylated ERK levels were correlated with time to tumor progression in the phase 2 trial conducted by Abou-Alfa et al.10 However, in patients with advanced HCC amenable only to systemic therapy, tumor tissue is generally not available. To our knowledge, no other clinical, biochemical, or radiologic parameters are available to date to guide the use of sorafenib in patients with advanced HCC. Interestingly, in the current study, we have demonstrated that the presence of extrahepatic involvement, in particular the presence of lung metastasis, was significantly associated with no clinical benefit with sorafenib treatment. This may be partly explained by the finding that the presence of lung metastasis is a poor prognostic factor for advanced HCC patients.14 It signifies high tumor load in advanced HCC patients that renders the patients refractory to sorafenib treatment. This provocative finding implies that sorafenib suggests the best therapeutic benefit in patients with locally advanced HCC. However, this important finding needs to be further validated in future studies. The identification of clinical factors predictive of clinical benefits is of particular importance in view of the high cost of sorafenib and potential toxicities associated with sorafenib.

In the SHARP study, approximately 52% of patients were reported to have grade 3 or 4 treatment-related toxicities, and the commonest adverse events were diarrhea (39%) and HFSR (21%).6 Similarly, the most commonly encountered toxicities in the present study were also diarrhea (67%) and HFSR (54%). Notably, there were higher incidences of malaise (55%), alopecia (37%), and rashes (29%) associated with sorafenib treatment in our study cohort. Moreover, greater than half of the patients encountered grade 1 or 2 hematologic toxicities. This might be partly attributed to the finding that most of the patients have some degree of pancytopenia associated with underlying cirrhosis at baseline. Nevertheless, there was no grade 3 or 4 leukopenia associated with sorafenib treatment in this study, and there was a far lower incidence of neutropenia in patients treated with sorafenib than in patients treated with single-agent doxorubicin or combination chemotherapy.15

Compared with the results of the reported studies,6, 16, 17 there was a notably high incidence of liver function derangement in our present patient cohort. These results were most likely related to a high proportion of patients with suboptimal baseline liver function as a result of more advanced cirrhosis in this study. The most common liver function derangement observed was the change in transaminase levels. Overall, there were only few cases of grade 4 liver dysfunctions, mainly in term of AST and bilirubin levels. Nonetheless, the majority of the liver derangements were transient and improved after stopping sorafenib. In fact, the mechanism of sorafenib-induced hepatotoxicity is still poorly elucidated and understood. It may be partly accounted for by the inhibition of UGT1A1 enzyme, which is responsible for hepatic bilirubin glucuronidation, by sorafenib.18 Furthermore, there are recent reports regarding the reactivation of hepatitis B infection in chronic hepatitis B carriers who received targeted therapy alone for the treatment of underlying malignancy.19, 20 Thus, it is possible that the administration of sorafenib will likewise lead to the reactivation of the underlying hepatitis B infection and result in worsening liver function, because most patients with advanced HCC in Asia are chronic hepatitis B carriers. Therefore, in view of these findings, vigilant liver function monitoring is advisable in patients with suboptimal baseline liver function on sorafenib treatment.

Thus far, based on the SHARP study that recruited mainly Child-Pugh A cirrhotic patients, sorafenib is only recommended for treatment of advanced HCC patients with underlying Child-Pugh A cirrhosis.21 However, in the current study, we demonstrated that there was no significant difference in clinical benefits from sorafenib in the treatment of Child-Pugh A or Child-Pugh B/C patients. Moreover, the results of the current study demonstrated that Child-Pugh B/C patients were not associated with higher incidence of grade 3 or 4 treatment-related toxicities from sorafenib when compared with Child-Pugh A patients. A recent retrospective subgroup analysis study conducted by Abou-Alfa et al18 did demonstrate similar findings to our present study. However, in that study, patients with Child-Pugh B cirrhosis fared worse than those with Child-Pugh A, with more frequent worsening of underlying cirrhosis. Therefore, more mature results are needed from other ongoing studies before generalized recommendation is made for sorafenib use in Child-Pugh B or C patients.

The major limitation of the current study is its nonrandomized design, and we could only compare the results with a historical cohort in our center. Moreover, quality of life is not included as part of the assessment in the current study. Nevertheless, the results from other studies6, 16 are only applicable to a highly selected Child-Pugh A population with good favorable parameters. Our results are more likely to reflect the results of “real life” situations in using sorafenib to treat advanced HCC patients than the aforementioned studies.

In conclusion, to our knowledge this is the first study in the literature that demonstrates good efficacy and reasonable tolerability of sorafenib in treating an advanced HCC patient subpopulation from a hepatitis B–endemic Asian population. Moreover, in this study, we have demonstrated that the presence of lung metastasis predicts poor response to sorafenib in advanced HCC patients.

Acknowledgements

We thank Prof. Raymond Liang and Richard Epstein for supporting multidisciplinary management of hepatocellular carcinoma.

Conflict of Interest Disclosures

Supported in part by the University of Hong Kong hapatocellular carcinoma research fund.

Ancillary