Purpose: Validated biomarkers of prognosis and response to drug have not been identified for patients with hepatocellular carcinoma (HCC). One of the objectives of the phase III, randomized, controlled Sorafenib HCC Assessment Randomized Protocol (SHARP) trial was to explore the ability of plasma biomarkers to predict prognosis and therapeutic efficacy.

Experimental Design: In SHARP, 602 patients with advanced HCC were randomized to receive either oral sorafenib 400 mg twice a day per os or matching placebo daily on a continuous basis. Ten plasma biomarkers implicated in the pathogenesis of HCC were measured in 491 patients at baseline and in 305 after 12 weeks of treatment. The candidate biomarkers were analyzed to identify correlates of prognosis or predictors of response to sorafenib.

Results: In both the entire patient population and the placebo cohort, baseline angiopoietin 2 (Ang2) and VEGF concentrations independently predicted survival. Clinical variables such as macroscopic vascular invasion, Eastern Cooperative Oncology Group (ECOG) performance status, and baseline α-fetoprotein and alkaline phosphatase concentrations also independently predicted survival in these groups. In the sorafenib cohort, trends toward enhanced survival benefit from sorafenib were observed in patients with high s-c-KIT or low hepatocyte growth factor concentration at baseline (P of interaction = 0.081 and 0.073, respectively).

Conclusions: The angiogenesis biomarkers Ang2 and VEGF were independent predictors of survival in patients with advanced HCC. In contrast, none of the biomarkers tested significantly predicted response to sorafenib. Clin Cancer Res; 18(8); 2290–300. ©2012 AACR.

Translational Relevance

Validated biomarkers of patient prognosis and response to treatment have not yet been identified in patients with hepatocellular carcinoma (HCC). We assessed whether baseline concentrations of 10 biomarkers and changes in their concentrations over 12 weeks could predict patient prognosis or response to treatment in the 602 patients enrolled in a registration phase III trial of sorafenib in patients with HCC. We found that the concentrations of two biomarkers, VEGF and Ang2, predicted patient survival, suggesting both may be included in prognostic staging systems for patients with HCC. We also found that concentrations of soluble c-KIT and hepatocyte growth factor tended to predict response to sorafenib. Further efforts are needed to identify biomarkers that can predict patient prognosis or response to treatment, thus allowing treatment to be individualized for patients with HCC.

Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related deaths worldwide and is associated with the second lowest 5-year survival rate of all tumor types (1). Moreover, HCC incidence and mortality rates appear to be increasing in the United States and other countries (2). Management of the disease, nonetheless, has improved over the last decade, largely as a result of advances in chemoembolization techniques and the advent of molecularly targeted therapy. The multikinase inhibitor sorafenib was shown in 2 randomized, double-blind, placebo-controlled trials to confer a significant survival benefit in patients with advanced HCC (3, 4), thereby establishing sorafenib as the standard systemic therapy for this indication (5, 6).

Biomarkers predicting patient prognosis or response to therapy may advance the potential of personalized medicine in cancer treatment (7). Previous studies have evaluated the correlation between baseline α-fetoprotein (AFP) concentration and patient outcomes (8–10). Additional studies have correlated various markers with survival in patients with HCC (11). They include the expression of epithelial cell adhesion molecule (EpCAM), a hepatic stem cell marker in tumor tissue (12–14); expression of the miR-26 miRNA precursor (15); and a prognostic gene signature in nontumor hepatic tissue (16). Because of the heterogeneity of HCC, however, the identification of biomarkers in this disease is somewhat complex. Although molecularly defined classes of HCC have not yet been linked to specific responses to treatment (17–19), signaling cascades involved in tumor proliferation and neoangiogenesis have been implicated in its pathogenesis. These cascades include several important kinases involved in tumor progression, several of which are pharmacologically relevant targets of sorafenib. The molecular targets of sorafenib include VEGFR-1, -2, and -3, platelet-derived growth factor receptor (PDGFR)-β, c-KIT, RET, FLT-3, and RAF (20, 21).

Previous investigations to identify prognostic biomarkers in patients with HCC have focused primarily on VEGF, angiopoietin 2 (Ang2), and hepatocyte growth factor (HGF; ref. 22–25), but these studies—conducted on tumor tissue (22–24) and hepatic vein (25) markers—have involved small numbers of patients. Biomarker evaluations in larger HCC patient populations, especially as part of randomized, placebo-controlled trials, may provide additional insight into the predictive and/or prognostic utility of specific markers.

One objective of the phase III Sorafenib HCC Assessment Randomized Protocol (SHARP) trial was to explore the ability of plasma biomarkers to predict patient prognosis and sorafenib efficacy. We therefore assayed plasma concentrations of 10 proteins that are either molecular targets of sorafenib [VEGF, soluble (s)-VEGFR-2 and -3, soluble c-KIT (s-c-KIT), and soluble Ras] or are known to interact with signaling pathways impacted by sorafenib and have been implicated in the pathogenesis of HCC on this basis [Ang2, basic fibroblast growth factor (bFGF), EGF, insulin-like growth factor (IGF)-2, and HGF] in patients who participated in the SHARP trial. To our knowledge, this is the largest study to date of these biomarkers in a randomized, placebo-controlled HCC trial population.

Patients and samples

The SHARP trial design has been described in detail (3). Eligible patients with advanced, measurable HCC, an Eastern Cooperative Oncology Group (ECOG) performance status (PS) ≤2 (26), and Child–Pugh class A (n = 602) status were randomized to sorafenib 400 mg twice a day (n = 299) or matching placebo (n = 303). The primary endpoints were overall survival (OS) and time to symptomatic progression (TTSP); secondary endpoints included time to progression (TTP) by independent radiological review, disease control rate (DCR), and safety.

Two 6-mL blood samples were collected at baseline (screening visit) and after 12 weeks of treatment, by venipuncture or through a Porta-a-Cath implantable venous access system, into a Vacutainer containing potassium EDTA. The blood samples were gently inverted and centrifuged within 10 to 15 minutes at 4°C for 10 minutes; if a refrigerated centrifuge was not available, the tubes were chilled on ice for 5 to 15 minutes and centrifuged in a standard centrifuge for 10 minutes. Plasma samples were frozen upright at ≤−70°C within 20 minutes of centrifugation and kept frozen until ready for shipment to the sponsor.

Biomarker assays

Plasma biomarker concentrations were measured by commercially available ELISA kits for Ang2, EGF, bFGF, VEGF, sVEGFR-2, sVEGFR-3, HGF, and s-c-KIT (R&D Systems; catalog numbers DANG20, DEG00, HSFB75, DVE00, DY349, DHG00, and DSCR00, respectively), IGF-2 (Diagnostic Systems Laboratories; catalog number DSL-10-2600), and all forms of circulating Ras (Oncogene Science Biomarker Group; catalog number 064900009), according to the manufacturers' instructions.

Outcomes

Clinical and outcome data used in these correlative analyses were obtained from the SHARP clinical trial database, with a May 12, 2006, cutoff date for TTP and a February 9, 2007, cutoff date for OS. To maximize statistical power and to provide the largest number of noncensored data points, we used an OS cutoff date for biomarker analysis that was approximately 4 months later than the October 17, 2006, cutoff date reported previously (3).

Statistical analysis

Statistical analyses were conducted with SAS and R software. For each biomarker, samples were dichotomized into 2 groups, as described later. Cox regression models and Kaplan–Meier analyses were used to assess the relationships of OS and TTP with baseline biomarker concentrations (for prognostic value) and changes in biomarker concentrations. Multivariate Cox proportional hazard models were used to evaluate the prognostic value for survival of these biomarkers, as well as of treatment group (sorafenib or placebo) and clinical variables previously identified as prognostic (3). The clinical variables included in the model are listed in relevant tables. Clinical variables in binned biomarker groups were compared using the F tests. The relationship between baseline biomarker levels and sorafenib treatment effect was evaluated using a Cox proportional hazards model with an interaction term. One-way ANOVA was used to compare changes in biomarker concentration from baseline to week 12 in the sorafenib and placebo groups.

All binary cutoff values were defined before data analysis. In the absence of clinical information on a cutoff value differentiating low and high baseline concentration of a plasma biomarker, we used the median concentrations: 11.3 ng/mL for s-c-KIT, 1,042.9 pg/mL for Ras, 8,653 pg/mL for sVEGFR-2, 39,587 pg/mL for sVEGFR-3, 6,061.1 pg/mL for Ang2, 7.5 pg/mL for bFGF, 30.4 pg/mL for EGF, and 797.7 ng/mL for IGF-2. If, however, clinical data substantiated use of a nonmedian cutoff value to differentiate low from high baseline biomarker concentration, then that nonmedian value was used. For example, serum HGF concentration above the 78th percentile (corresponding to 1.0 ng/mL) was associated with poor survival in 55 patients with inoperable HCC (27); we therefore used the 75th percentile (3,279.1 pg/mL) as the cutoff value distinguishing low and high plasma HGF concentration. (We did not use the absolute value of 1.0 ng/mL as a cutoff value due to possible differences in serum and plasma concentrations). Although a phase III trial of sorafenib in patients with advanced renal cell carcinoma showed a trend toward greater sorafenib benefit in patients with VEGF level above the median, this relationship became significant when the 75th percentile was used as the cutoff value (28). We therefore used the 75th percentile (101.9 pg/mL) to differentiate a low from a high baseline VEGF concentration. To analyze the correlation between change in biomarker concentration and outcome, the median percent changes among all patients was used as the cutoff value to differentiate low from high changes.

Because the SHARP trial biomarker analyses were exploratory and hypothesis generating, no P value corrections for multiple testing were conducted.

Populations of patients evaluated for biomarkers

A total of 602 patients were randomized in the SHARP trial, 299 to sorafenib and 303 to placebo (3). Trial centers submitted baseline plasma samples from 499 patients; 12-week samples were analyzed only if a baseline sample was available from the same patient. Ultimately, usable plasma samples were received from 491 patients (81.6%) at baseline and from 305 (50.7%) at 12 weeks. All usable plasma samples were assayed for all 10 proteins, as plasma volume allowed; because plasma volumes varied, the number of patients with data available for each biomarker ranged from 485 to 491 at baseline and from 274 to 305 at 12 weeks.

Baseline demographic and disease characteristics of patients in the biomarker subpopulations were similar to those in the overall SHARP population, as were the clinical benefits of sorafenib. In the SHARP biomarker population, OS in the sorafenib and placebo groups was 10.8 and 8.5 months, respectively [HR, 0.72; 95% confidence interval (CI), 0.58–0.90], and TTP was 5.3 and 3.0 months (HR, 0.60; 95% CI, 0.46–0.79), respectively. In comparison, OS of the sorafenib and placebo groups in the overall SHARP population was 10.7 and 7.9 months (HR, 0.69; 95% CI, 0.55–0.87), respectively, and TTP was 5.5 and 2.8 months (HR, 0.58; 95% CI, 0.45–0.74), respectively.

Prognostic value of plasma biomarkers for all patients

We first analyzed the prognostic value of plasma biomarkers in all randomized patients in the SHARP trial. Both baseline Ang2 and baseline VEGF concentrations correlated with survival (Fig. 1A and B). The median survival of patients with low and high baseline Ang2 concentrations was 14.1 and 6.3 months, respectively, whereas the median survival of patients with low and high baseline VEGF concentrations was 10.6 and 6.2 months, respectively. A multivariate analysis that included all 10 baseline plasma biomarkers, treatment group, and the predictors of survival previously identified in patients with advanced HCC (3) showed that, among the entire SHARP population, both baseline Ang2 and VEGF retained independent prognostic value, along with treatment group, ECOG PS, macrovascular invasion, and baseline plasma levels of AFP and alkaline phosphatase (Table 1).

Figure 1.

Univariate analyses of baseline plasma biomarkers as prognostic factors in advanced HCC. Ang2 (A) and VEGF (B) versus OS in the full cohort of patients. C, Ang2, (D) VEGF, (E) HGF, and (F) IGF-2 versus OS in the placebo cohort.

Figure 1.

Univariate analyses of baseline plasma biomarkers as prognostic factors in advanced HCC. Ang2 (A) and VEGF (B) versus OS in the full cohort of patients. C, Ang2, (D) VEGF, (E) HGF, and (F) IGF-2 versus OS in the placebo cohort.

Close modal
Table 1.

Multivariate analyses of the sorafenib, placebo, and all-patient cohorts to identify factors independently prognostic for OS in patients with HCC

Placebo cohortSorafenib cohortAll-patient cohort
Baseline factorP valueHR (95% CI)P valueHR (95% CI)P valueHR (95% CI)
Treatment (sorafenib vs. placebo) NA — NA — 0.041 0.78 (0.61–0.99) 
ECOG PS (0 vs. >0) NS — NS — 0.016 1.36 (1.06–1.75) 
AFP (≤ vs. >200 ng/mL) 0.007 1.58 (1.13–2.21) 0.015 1.57 (1.09–2.26) 0.001 1.49 (1.17–1.89) 
MVI (present vs. absent) 0.0003 1.88 (1.34–2.66) 0.005 1.74 (1.19–2.55) <0.0001 1.81 (1.43–2.32) 
EHS (present vs. absent) NS — 0.016 1.60 (1.09–2.34) NS — 
Alkaline phosphatase (≤ vs. >median) 0.013 1.56 (1.10–2.22) 0.002 1.82 (1.26–2.63) 0.0003 1.59 (1.24–2.04) 
s-c-KIT (≤ vs. >11.3 ng/mL) NS — 0.004 0.56 (0.38–0.83) NS — 
HGF (≤ vs. >3,279.1 pg/mL) NS — 0.017 1.69 (1.10–2.60) NS — 
VEGF (≤ vs. >101.9 pg/mL) 0.002 1.97 (1.28–3.04) NS — 0.015 1.48 (1.08–2.03) 
Ang2 (≤ vs. >6,043.5 pg/mL) 0.002 1.84 (1.25–2.70) 0.034 1.59 (1.04–2.43) 0.001 1.58 (1.20–2.07) 
Placebo cohortSorafenib cohortAll-patient cohort
Baseline factorP valueHR (95% CI)P valueHR (95% CI)P valueHR (95% CI)
Treatment (sorafenib vs. placebo) NA — NA — 0.041 0.78 (0.61–0.99) 
ECOG PS (0 vs. >0) NS — NS — 0.016 1.36 (1.06–1.75) 
AFP (≤ vs. >200 ng/mL) 0.007 1.58 (1.13–2.21) 0.015 1.57 (1.09–2.26) 0.001 1.49 (1.17–1.89) 
MVI (present vs. absent) 0.0003 1.88 (1.34–2.66) 0.005 1.74 (1.19–2.55) <0.0001 1.81 (1.43–2.32) 
EHS (present vs. absent) NS — 0.016 1.60 (1.09–2.34) NS — 
Alkaline phosphatase (≤ vs. >median) 0.013 1.56 (1.10–2.22) 0.002 1.82 (1.26–2.63) 0.0003 1.59 (1.24–2.04) 
s-c-KIT (≤ vs. >11.3 ng/mL) NS — 0.004 0.56 (0.38–0.83) NS — 
HGF (≤ vs. >3,279.1 pg/mL) NS — 0.017 1.69 (1.10–2.60) NS — 
VEGF (≤ vs. >101.9 pg/mL) 0.002 1.97 (1.28–3.04) NS — 0.015 1.48 (1.08–2.03) 
Ang2 (≤ vs. >6,043.5 pg/mL) 0.002 1.84 (1.25–2.70) 0.034 1.59 (1.04–2.43) 0.001 1.58 (1.20–2.07) 

NOTE: Analyses included the following baseline variables: all biomarker values, ECOG PS, AFP, macroscopic vascular invasion (MVI), extrahepatic spread (EHS), albumin score, alkaline phosphatase, bilirubin score, prothrombin time, presence of ascites, and, for the all-patient cohort analysis, treatment group. Factors with P < 0.05 in one or more of the analyses are shown.

Abbreviations: NA, not applicable; NS, not significant (P > 0.05).

Biomarkers prognostic in the placebo group

Univariate analyses of the potential prognostic value of the 10 candidate biomarkers in the placebo group alone showed that low baseline concentrations of Ang2, VEGF, and HGF, and high baseline concentrations of IGF-2, correlated with better OS (Fig. 1C–F). Low baseline concentrations of Ang2 also correlated with longer TTP (HR, 1.52; P = 0.016; data not shown), with Ang2 being the only biomarker prognostic for both OS and TTP. Baseline levels of bFGF and the other biomarkers assayed did not correlate with prognosis among patients in the placebo group (data not shown). Baseline concentrations of HGF, VEGF, s-c-KIT, Ang2, and IGF-2 correlated with other clinical/demographic variables associated with poor outcome in advanced HCC, including an ECOG PS of 1 or 2; macroscopic vascular invasion and/or extrahepatic spread; and concentrations of AFP, albumin, alkaline phosphatase, and bilirubin (Table 2). In addition, multivariate analysis—which included all 10 biomarkers plus clinical factors previously found to be prognostic in patients with advanced HCC (3)—showed that baseline Ang2 and VEGF concentrations were independently prognostic for OS (P = 0.002 each).

Table 2.

Univariate analyses of baseline biomarker concentrations and demographic/clinical variables

HGFVEGFs-c-KITAng2IGF-2
Demographic or clinical variableMean level, pg/mLP valueMean level, pg/mLP valueMean level, ng/mLP valueMean level, pg/mLP valueMean level, pg/mLP value
Sex           
 Male 2,272.2 NS 94.2 NS 12.0 NS 7,672.5 NS 840.7 <0.0001 
 Female 2,754.2  93.2  12.4  8,175.0  1,374.7  
Age           
 <65 2,768.5 NS 106.0 NS 12.5 NS 7,850.3 NS 972.7 0.0182 
 ≥65 2,770.6  86.1  11.9  7,668.6  873.6  
ECOG PS           
 0 2,778.6 NS 82.8 0.047 13.0 0.003 6,608.8 <0.0001 935.4 NS 
 1 or 2 2,760.3  106.0  11.1  8,917.6  889.9  
MVI and/or EHS           
 Absent 2,635.4 NS 81.9 NS 12.6 NS 6,800.0 0.018 885.2 NS 
 Present 2,830.9  99.5  11.9  8,162.2  925.6  
AFP           
 ≤Median 2,549.5 0.011 97.4 NS 12.1 NS 6,887.8 0.0015 947.2 NS 
 >Median 2,989.0  90.7  12.1  8,590.7  879.2  
MVI           
 No 2,570.4 0.002 94.4 NS 11.9 NS 6,889.6 <0.0001 960.8 0.001 
 Yes 3,096.4  93.4  12.4  9,133.1  835.1  
Albumin           
 ≤Median 3,145.5 <0.001 98.4 NS 12.8 0.032 9,273.2 <0.0001 754.6 <0.0001 
 >Median 2,398.8  89.7  11.4  6,189.2  1,073.0  
Alkaline phosphatase           
 ≤Median 2,472.1 <0.001 78.8 0.004 11.8 NS 6,305.7 <0.0001 929.6 NS 
 >Median 3,071.2  109.6  12.4  9,195.0  896.2  
Total bilirubin           
 ≤Median 2,412.8 <0.001 107.3 0.016 10.7 <0.0001 6,857.3 0.0003 1,056.9 <0.0001 
 >Median 3,122.2  81.1  13.5  8,598.9  771.7  
HGFVEGFs-c-KITAng2IGF-2
Demographic or clinical variableMean level, pg/mLP valueMean level, pg/mLP valueMean level, ng/mLP valueMean level, pg/mLP valueMean level, pg/mLP value
Sex           
 Male 2,272.2 NS 94.2 NS 12.0 NS 7,672.5 NS 840.7 <0.0001 
 Female 2,754.2  93.2  12.4  8,175.0  1,374.7  
Age           
 <65 2,768.5 NS 106.0 NS 12.5 NS 7,850.3 NS 972.7 0.0182 
 ≥65 2,770.6  86.1  11.9  7,668.6  873.6  
ECOG PS           
 0 2,778.6 NS 82.8 0.047 13.0 0.003 6,608.8 <0.0001 935.4 NS 
 1 or 2 2,760.3  106.0  11.1  8,917.6  889.9  
MVI and/or EHS           
 Absent 2,635.4 NS 81.9 NS 12.6 NS 6,800.0 0.018 885.2 NS 
 Present 2,830.9  99.5  11.9  8,162.2  925.6  
AFP           
 ≤Median 2,549.5 0.011 97.4 NS 12.1 NS 6,887.8 0.0015 947.2 NS 
 >Median 2,989.0  90.7  12.1  8,590.7  879.2  
MVI           
 No 2,570.4 0.002 94.4 NS 11.9 NS 6,889.6 <0.0001 960.8 0.001 
 Yes 3,096.4  93.4  12.4  9,133.1  835.1  
Albumin           
 ≤Median 3,145.5 <0.001 98.4 NS 12.8 0.032 9,273.2 <0.0001 754.6 <0.0001 
 >Median 2,398.8  89.7  11.4  6,189.2  1,073.0  
Alkaline phosphatase           
 ≤Median 2,472.1 <0.001 78.8 0.004 11.8 NS 6,305.7 <0.0001 929.6 NS 
 >Median 3,071.2  109.6  12.4  9,195.0  896.2  
Total bilirubin           
 ≤Median 2,412.8 <0.001 107.3 0.016 10.7 <0.0001 6,857.3 0.0003 1,056.9 <0.0001 
 >Median 3,122.2  81.1  13.5  8,598.9  771.7  

Abbreviations: EHS, extrahepatic spread; MVI, macroscopic vascular invasion; NS, not significant (P > 0.05).

Biomarkers prognostic in the sorafenib group

Because sorafenib is the current standard of care worldwide for patients with advanced HCC (3–5, 28–30), we analyzed the correlation of clinical factors and biomarkers with outcome in the sorafenib group of the SHARP study. Multivariate analysis showed that s-c-KIT, HGF, and Ang2 concentrations were independent prognostic factors for OS in patients treated with sorafenib. Although VEGF concentration was prognostic for patients in the all-patient and placebo cohorts, it was not prognostic for patients treated with sorafenib.

To determine whether plasma biomarkers could predict response to sorafenib, we analyzed the interaction between each baseline biomarker and sorafenib treatment effect. We found that patients with a high baseline s-c-KIT concentration tended to show greater improvements in OS (P of interaction = 0.081; Fig. 2A and B) and TTP (P of interaction = 0.052; Fig. 3A and B) in response to sorafenib than in those with a low s-c-KIT level. Conversely, patients with low baseline HGF tended to derive greater benefit from sorafenib in both OS (P of interaction = 0.073; Fig. 2C and D) and TTP (P of interaction = 0.396; data not shown) than those with high HGF concentration. In addition, patients with high baseline bFGF tended to show greater sorafenib-associated improvement in TTP (P of interaction = 0.078; Fig. 3C and D) than those with low bFGF level; however, a similar association was not observed for bFGF and OS benefit, where both high and low bFGF groups benefited equally from sorafenib treatment (P of interaction = 0.46; data not shown). Although baseline Ang2 concentrations correlated with OS in multivariate analysis of the sorafenib cohort alone, the biomarker treatment interaction analysis did not correlate with sorafenib-associated survival benefit (P of interaction = 0.80; data not shown).

Figure 2.

Analysis of baseline biomarkers as predictive factors for sorafenib benefit (OS). Low s-c-KIT (A) and high s-c-KIT (B), P value for biomarker treatment interaction = 0.081. C, low HGF and (D) high HGF, P value for biomarker treatment interaction = 0.073.

Figure 2.

Analysis of baseline biomarkers as predictive factors for sorafenib benefit (OS). Low s-c-KIT (A) and high s-c-KIT (B), P value for biomarker treatment interaction = 0.081. C, low HGF and (D) high HGF, P value for biomarker treatment interaction = 0.073.

Close modal
Figure 3.

Analysis of baseline biomarkers as predictive factors for sorafenib benefit (TTP). Low s-c-KIT (A) and high s-c-KIT (B), P value for biomarker treatment interaction = 0.052. Low bFGF (C) and high bFGF (D), P value for biomarker treatment interaction = 0.078.

Figure 3.

Analysis of baseline biomarkers as predictive factors for sorafenib benefit (TTP). Low s-c-KIT (A) and high s-c-KIT (B), P value for biomarker treatment interaction = 0.052. Low bFGF (C) and high bFGF (D), P value for biomarker treatment interaction = 0.078.

Close modal

Treatment-induced changes in plasma biomarker concentrations and correlation with outcome

We found that the change from baseline to week 12 in mean plasma concentration of 8 biomarkers differed significantly between the sorafenib and placebo groups (Fig. 4). For example, mean plasma s-c-KIT concentration decreased significantly in the sorafenib group but was essentially unchanged in the placebo group (P < 0.0001). Mean plasma HGF decreased in the sorafenib group but increased in the placebo group (P < 0.0001). In the sorafenib group, mean plasma concentration of VEGF increased significantly (P = 0.010), and mean concentrations of sVEGFR-2 (P < 0.0001) and sVEGFR-3 (P < 0.0001) decreased significantly, compared with levels in the placebo group. Interestingly, the mean concentration of Ang2—a biomarker we found to be independently prognostic for survival—increased in the placebo group but did not change significantly in the sorafenib group (P < 0.0001). Mean levels of bFGF and the other biomarkers tested did not change differently between the sorafenib and placebo groups.

Figure 4.

Analysis of the change in biomarker levels during treatment. Mean plasma concentrations of (A) s-c-KIT, (B) HGF, (C) VEGF, (D) sVEGFR-2, (E) sVEGFR-3, (F) Ras, (G) Ang2, and (H) bFGF at baseline (black bars) and at week 12 (gray bars). The P values compare the changes from baseline to week 12 in each biomarker concentration between the sorafenib and placebo groups by one-way ANOVA.

Figure 4.

Analysis of the change in biomarker levels during treatment. Mean plasma concentrations of (A) s-c-KIT, (B) HGF, (C) VEGF, (D) sVEGFR-2, (E) sVEGFR-3, (F) Ras, (G) Ang2, and (H) bFGF at baseline (black bars) and at week 12 (gray bars). The P values compare the changes from baseline to week 12 in each biomarker concentration between the sorafenib and placebo groups by one-way ANOVA.

Close modal

Cox regression models and Kaplan–Meier analyses were conducted to examine the relationships between changes in biomarker concentrations and outcome. Ang2 increases of at least 5.1% (the median change in Ang2) were associated with shorter OS and TTP (data not shown) in both the sorafenib (OS, P < 0.0001; TTP, P = 0.0002) and placebo (OS, P < 0.0001; TTP, P < 0.0001) cohorts, reflecting our finding that Ang2 is a biomarker of poor prognosis in patients with HCC. A decrease in HGF level of >2.7% (the median change in HGF) was associated with longer TTP (P = 0.042) but not longer OS (P = 0.0521) among sorafenib-treated patients and with both longer TTP (P < 0.0001) and OS (P < 0.000001) among patients who received placebo. A decrease in mean IGF-2 level of >11.2% (the median change in IGF-2) was associated with shorter OS (P = 0.005) and a trend toward shorter TTP (P = 0.075) among sorafenib-treated patients as well as shorter OS (P < 0.0001) and shorter TTP (P = 0.009) among patients who received placebo. No associations between change in biomarker concentration and outcome (either OS or TTP) were identified for bFGF or the other biomarker candidates tested (P > 0.05 for all).

This study represents the largest effort to date to identify biomarkers of prognosis and response to sorafenib in patients with advanced HCC. This study was conducted in the setting of the phase III SHARP trial (3), which evaluated the efficacy and safety of sorafenib in patients with advanced HCC. We found that a number of plasma biomarkers—including Ang2, VEGF, HGF, and IGF-2—were predictors of prognosis in patients with advanced HCC, but none of the plasma markers tested significantly predicted response to sorafenib.

The most important finding of this study is the identification of Ang2 and VEGF as strong, independent predictors of survival in patients with HCC. Ang2 and VEGF are key signaling elements that drive angiogenesis, thereby enabling HCC growth and metastasis (31). To our knowledge, this study is the first to suggest that high plasma Ang2 concentrations at baseline are indicative of poor prognosis in patients with advanced HCC, suggesting that elevated levels of this angiogenic factor may be associated with more aggressive disease. Ang2 concentrations increased during treatment in the placebo group, suggesting poor outcome related to disease progression in this cohort. In contrast, Ang2 levels appear to be held constant during treatment with sorafenib, perhaps reflecting the more favorable outcome in this group. Furthermore, increases in Ang2 during treatment were associated with poorer outcomes in both groups, suggesting that measurements of Ang2 may have value in disease monitoring during treatment.

As elevated levels of VEGF at baseline indicate poor prognosis in patients with advanced HCC, a result consistent with previous findings (32–34), the increase in VEGF concentration observed after sorafenib treatment is at first glance counterintuitive, given the known survival advantages of sorafenib in patients with advanced HCC (3). However, treatment-induced increases in plasma concentration of VEGF (along with decreases in sVEGFR-2) have been consistently observed in other trials of sorafenib (35, 36) and with other agents inhibiting VEGFR-2 in HCC (37, 38) and other tumor types (39–41). Treatment with the anti-VEGF antibody bevacizumab has yielded mixed results, with increases in VEGF observed in some studies (40, 42) and decreases in VEGFR noted in others, including HCC (43, 44). Increases in VEGF, and associated decreases in VEGFR-2, have also been observed in nontumor-bearing mice after treatment with a VEGFR-2 inhibitor (45), suggesting that at least part of the increase in VEGF observed in humans is tumor independent. In the current study, the change in VEGF level observed during treatment did not correlate with outcome. Thus, these treatment-induced increases in VEGF are likely to be (at least in part) tumor independent and may not adversely affect the tumor due to efficient blockage of VEGFR signaling by sorafenib.

The role of the Ang-Tie2 pathway in oncogenesis has been reviewed recently (46). Increased expression of Ang2, particularly in conjunction with high VEGF-A concentration, correlated with poor outcomes in patients with breast cancer (47) and those with non–small cell lung cancer (NSCLC; ref. 48), as well as those with advanced HCC (49). Our biomarker analysis suggests that both molecules are independent predictors of survival in patients with HCC and provides a basis for novel opportunities for combination therapy in these patients.

Elevated HGF concentration was also identified as indicative of poor prognosis in the present study, although HGF did not retain significance in multivariate modeling. Commensurate with this finding, mean HGF levels decreased during treatment with sorafenib (and not in the placebo cohort), perhaps reflecting the more favorable outcome of the sorafenib group. Again, consistently, patients in either treatment group exhibiting HGF decreases greater than the median experienced better outcomes (longer OS and/or TTP). Taken together, these data suggest that HGF levels directly reflect HCC disease status, with low levels indicating favorable prognosis and decreasing levels suggesting disease improvement.

A few biomarkers predicting drug response have been confirmed in oncology, including HER2 expression and response to trastuzumab in breast and gastric cancers (50) and KRAS mutations and responses to cetuximab (51) and panitumumab (52) in colon cancer. The number is increasing steadily, including the recent approvals of vemurafenib for patients with BRAF V600E mutation-positive metastatic melanoma (53) and crizotinib for patients with NSCLC and anaplastic lymphoma kinase (ALK) rearrangement, as shown by fluorescence in situ hybridization tests (54). These predictive biomarkers have thus far been identified in tumor tissue rather than plasma samples. In the present study, we found that baseline plasma concentrations of s-c-KIT and HGF were independent predictors of survival in patients receiving sorafenib but showed only a nonsignificant trend as predictors of response to sorafenib treatment. The clinical significance of these results is uncertain as the role of s-c-KIT in the pathogenesis of HCC has not been consistently showed. Of further note are the results for HGF, a ligand that signals through the receptor tyrosine kinase c-MET. The HGF-MET cascade is relevant in hepatocarcinogenesis, and MET activation has been associated with poor outcome (55). In preclinical models, greater concentrations of sorafenib were required to inhibit the proliferation of HCC cells cultured with HGF than those without (56). Our clinical results may reflect this finding, in that patients with elevated HGF levels at baseline showed a trend toward deriving less benefit from sorafenib than those with low levels. Studies in NSCLC may explain this phenomenon, suggesting that HGF may be involved in conferring resistant to treatment with receptor tyrosine kinase inhibitors (57).

The clinical ramifications of the findings from this exploratory biomarker analysis of a large, randomized, placebo-controlled cohort of patients are 3-fold. First, we found that plasma Ang2 and VEGF concentrations, in addition to AFP concentration and other clinical parameters, are independent predictors of survival and should be considered prognostic biomarkers in patients with advanced HCC. Second, these prognostic biomarkers may prove valuable for the stratification of patients with advanced HCC before randomization in clinical trials. Finally, although trends of interest were identified in plasma s-c-KIT and HGF levels as predictive markers, none of the plasma biomarkers tested reached statistical significance in predicting response to sorafenib. Before any of these biomarkers can be used clinically as surrogate markers of efficacy or response to sorafenib, further investigations are needed to confirm and validate their predictive and/or prognostic value.

The following principal investigators (listed alphabetically by country) enrolled patients in the SHARP trial: Argentina: M.G. Pallota, J.J. Zarba; Australia: M. Boyer, S. Riordan, A. Strickland, N. Tebbutt, B. Thomson; Belgium: I. Borbath, J. De Greve, J.-L. Van Laethem, W. Van Steenbergen, H. Van Vlierberghe; Brazil: C. Barrios, A. Cosme de Oliveira; Bulgaria: I. Kotzev, D. Takov, K. Tchernev; Canada: K. Burak, M. Ma, P. Metrakos, C. Olweny, M. Sherman; Chile: C. Gamargo Garate, J. Martinez-Castillo; France: M. Beaugrand, J. Bennouna, J.-F. Blanc, J.-P. Bronowicki, F. Degos, S. Dominguez, J.-D. Grange, P. Hillon, J.-L. Raoul, J.-F. Seitz; Germany: H. Blum, P. Buggisch, W. Caspary, M. Dollinger, P.R. Galle, G. Gerken, B. Göke, M. Gregor, T. Greten, D. Häussinger, P. Hilgard, J. Scherübl, M. Scheulen, R. Schmid, U. Spengler, R. Wiest, S. Zeuzem; Greece: C. Arvanitakis, G. Germanidis, I. Katsos; Israel: A. Figer, S. Stemmer; Italy: D. Amadori, L. Bolondi, F. Cognetti, A. Craxi, F. Farinati, C. Gridelli, A. Martoni, V. Mazzaferro, C. Porta, S. Ricci, A. Sangiovanni, A. Santoro, F. Trevisani; Mexico: L.E. Cisnero Garza; New Zealand: E. Gane, A. O'Donnell; Peru: J. Leon, A. Lozano; Poland: J. Jassem, G. Rydzewska, A. Szawlowski, P. Tomczak; Romania: F. Badulescu, L. Miron; Russia: V. Kubyshkin; Spain: J. Bruix, A. Forner, J. Bustamante Schneider, M. Diago, J.L. Montero Alvarez, S. Pascual, L. Ruíz del Arbol, B. Sangro, R. Solá, J. Tabernero; Switzerland: B. Muellhaupt, A. Roth; United Kingdom: T.R. Jeffry Evans, S. Falk, T. Meyer, H. Reeves, P. Ross; United States: A. Befeler, T. Boyer, C. Britten, T. Byrne, G. Garcia-Tsao, P. Gold, A. Goldenberg, D. Heuman, P. Kennedy, A. Koch, J.M. Llovet, J. Marrero, M. Schilsky, J. Schwartz, M. Schwartz.

J.M. Llovet and J. Bruix have participated in advisory activities for Bayer HealthCare Pharmaceuticals and Onyx Pharmaceuticals. No potential conflicts of interest were disclosed by the other authors.

The authors thank Scott M. Wilhelm, PhD, of Bayer HealthCare Pharmaceuticals (Montville, NJ) and BelMed Professional Resources (New Rochelle, NY) for editorial support.

The study was supported by Bayer HealthCare Pharmaceuticals and Onyx Pharmaceuticals. J.M. Llovet has grants from NIH—NIDDK 1R01DK076986-01, FP7-2010-Health European Commission grant HEPTROMIC: 259744-2; NIH (Spain) grant I+D Program (SAF-2007-61898; SAF-2010-16055); and Samuel Waxman Cancer Research Foundation. J. Bruix has grants from the Instituto de Salud Carlos III (PI 08/0146 and PI 11/01830). J.M. Llovet and J. Bruix have been provided research support for Bayer HealthCare Pharmaceuticals.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Jemal
A
,
Siegel
R
,
Xu
J
,
Ward
E
. 
Cancer statistics, 2010
.
CA Cancer J Clin
2010
;
60
:
277
300
.
2.
Altekruse
SF
,
McGlynn
KA
,
Reichman
ME
. 
Hepatocellular carcinoma incidence, mortality, and survival trends in the United States from 1975 to 2005
.
J Clin Oncol
2009
;
27
:
1485
91
.
3.
Llovet
JM
,
Ricci
S
,
Mazzaferro
V
,
Hilgard
P
,
Gane
E
,
Blanc
JF
SHARP Investigators Study Group
. 
Sorafenib in advanced hepatocellular carcinoma
.
N Engl J Med
2008
;
359
:
378
90
.
4.
Cheng
AL
,
Kang
YK
,
Chen
Z
,
Tsao
CJ
,
Qin
S
,
Kim
JS
, et al
Efficacy and safety of sorafenib in patients in the Asia-Pacific region with advanced hepatocellular carcinoma: a phase III randomised, double-blind, placebo-controlled trial
.
Lancet Oncol
2009
;
10
:
25
34
.
5.
Llovet
JM
,
Di Bisceglie
AM
,
Bruix
J
,
Kramer
BS
,
Lencioni
R
,
Zhu
AX
Panel of Experts in HCC-Design Clinical Trials
. 
Design and endpoints of clinical trials in hepatocellular carcinoma
.
J Natl Cancer Inst
2008
;
100
:
698
711
.
6.
Bruix
J
,
Sherman
M;
American Association for the Study of Liver Diseases
. 
Management of hepatocellular carcinoma: an update
.
Hepatology
2011
;
53
:
1020
2
.
7.
Sawyers
CL
. 
The cancer biomarker problem
.
Nature
2008
;
452
:
548
52
.
8.
Okuda
K
,
Tanaka
M
,
Kanazawa
N
,
Nagashima
J
,
Satomura
S
,
Kinoshita
H
, et al
Evaluation of curability and prediction of prognosis after surgical treatment for hepatocellular carcinoma by lens culinaris agglutinin-reactive alpha-fetoprotein
.
Int J Oncol
1999
;
14
:
256
71
.
9.
Riaz
A
,
Ryu
RK
,
Kulik
LM
,
Mulcahy
MF
,
Lewandowski
RJ
,
Minocha
J
, et al
Alpha-fetoprotein response after locoregional therapy for hepatocellular carcinoma: oncologic marker of radiologic response, progression, and survival
.
J Clin Oncol
2009
;
27
:
5734
42
.
10.
Tamura
Y
,
Igarashi
M
,
Kawai
H
,
Suda
T
,
Satomura
S
,
Aoyagi
Y
. 
Clinical advantage of highly sensitive on-chip immunoassay for fucosylated fraction of alpha-fetoprotein in patients with hepatocellular carcinoma
.
Dig Dis Sci
2010
;
55
:
3576
83
.
11.
Villanueva
A
,
Hoshida
Y
,
Toffanin
S
,
Lachenmayer
A
,
Alsinet
C
,
Savic
R
, et al
New strategies in hepatocellular carcinoma: genomic prognostic markers
.
Clin Cancer Res
2010
;
16
:
4688
94
.
12.
Terris
B
,
Cavard
C
,
Perret
C
. 
EpCAM, a new marker for cancer stem cells in hepatocellular carcinoma
.
J Hepatol
2010
;
52
:
280
1
.
13.
Yamashita
T
,
Forgues
M
,
Wang
W
,
Kim
JW
,
Ye
Q
,
Jia
H
, et al
EpCAM and alpha-fetoprotein expression defines novel prognostic subtypes of hepatocellular carcinoma
.
Cancer Res
2008
;
68
:
1451
61
.
14.
Yamashita
T
,
Ji
J
,
Budhu
A
,
Forgues
M
,
Yang
W
,
Wang
HY
, et al
EpCAM-positive hepatocellular carcinoma cells are tumor-initiating cells with stem/progenitor cell features
.
Gastroenterology
2009
;
136
:
1012
24
.
15.
Ji
J
,
Shi
J
,
Budhu
A
,
Yu
Z
,
Forgues
M
,
Roessler
S
, et al
MicroRNA expression, survival, and response to interferon in liver cancer
.
N Engl J Med
2009
;
361
:
1437
47
.
16.
Hoshida
Y
,
Villanueva
A
,
Kobayashi
M
,
Peix
J
,
Chiang
DY
,
Camargo
A
, et al
Gene expression in fixed tissues and outcome in hepatocellular carcinoma
.
N Engl J Med
2008
;
359
:
1995
2004
.
17.
Hoshida
Y
,
Toffanin
S
,
Lachenmayer
A
,
Villanueva
A
,
Minguez
B
,
Llovet
JM
. 
Molecular classification and novel targets in hepatocellular carcinoma: recent advancements
.
Semin Liver Dis
2010
;
30
:
35
51
.
18.
Llovet
JM
,
Bruix
J
. 
Molecular targeted therapies in hepatocellular carcinoma
.
Hepatology
2008
;
48
:
1312
27
.
19.
Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet 2012, in press. DOI:10.1016/S0140-6736(11)61347-0.
20.
Wilhelm
SM
,
Carter
C
,
Tang
L
,
Wilkie
D
,
McNabola
A
,
Rong
H
, et al
BAY 43-9006 exhibits broad spectrum oral antitumor activity and targets the RAF/MEK/ERK pathway and receptor tyrosine kinases involved in tumor progression and angiogenesis
.
Cancer Res
2004
;
64
:
7099
109
.
21.
Wilhelm
SM
,
Adnane
L
,
Newell
P
,
Villanueva
A
,
Llovet
JM
,
Lynch
M
. 
Preclinical overview of sorafenib, a multikinase inhibitor that targets both Raf and VEGF and PDGF receptor tyrosine kinase signaling
.
Mol Cancer Ther
2008
;
7
:
3129
40
.
22.
Mitsuhashi
N
,
Shimizu
H
,
Ohtsuka
M
,
Wakabayashi
Y
,
Ito
H
,
Kimura
F
, et al
Angiopoietins and Tie-2 expression in angiogenesis and proliferation of human hepatocellular carcinoma
.
Hepatology
2003
;
37
:
1105
13
.
23.
Torimura
T
,
Sata
M
,
Ueno
T
,
Kin
M
,
Tsuji
R
,
Suzaku
K
, et al
Increased expression of vascular endothelial growth factor is associated with tumor progression in hepatocellular carcinoma
.
Hum Pathol
1998
;
29
:
986
91
.
24.
Mizuguchi
T
,
Nagayama
M
,
Meguro
M
,
Shibata
T
,
Kaji
S
,
Nobuoka
T
, et al
Prognostic impact of surgical complications and preoperative serum hepatocyte growth factor in hepatocellular carcinoma patients after initial hepatectomy
.
J Gastrointest Surg
2009
;
13
:
325
33
.
25.
Kuboki
S
,
Shimizu
H
,
Mitsuhashi
N
,
Kusashio
K
,
Kimura
F
,
Yoshidome
H
, et al
Angiopoietin-2 levels in the hepatic vein as a useful predictor of tumor invasiveness and prognosis in human hepatocellular carcinoma
.
J Gastroenterol Hepatol
2008
;
23
:
e157
64
.
26.
Oken
MM
,
Creech
RH
,
Tormey
DC
,
Horton
J
,
Davis
TE
,
McFadden
ET
, et al
Toxicity and response criteria of the Eastern Cooperative Oncology Group
.
Am J Clin Oncol
1982
;
5
:
649
55
.
27.
Vejchapipat
P
,
Tangkijvanich
P
,
Theamboonlers
A
,
Chongsrisawat
V
,
Chittmittrapap
S
,
Poovorawan
Y
. 
Association between serum hepatocyte growth factor and survival in untreated hepatocellular carcinoma
.
J Gastroenterol
2004
;
39
:
1182
8
.
28.
Omata
M
,
Lesmana
LA
,
Tateishi
R
,
Chen
PJ
,
Lim
SM
,
Yoshida
H
, et al
Asian Pacific Association for the Study of the Liver consensus recommendations on hepatocellular carcinoma
.
Hepatol Int
2010
;
4
:
439
74
.
29.
Poon
D
,
Anderson
BO
,
Chen
LT
,
Tanaka
K
,
Lau
WY
,
Van Cutsem
E
Asian Oncology Summit
. 
Management of hepatocellular carcinoma in Asia: consensus statement from the Asian Oncology Summit 2009
.
Lancet Oncol
2009
;
10
:
1111
8
.
30.
Benson
AB
 III
,
Abrams
TA
,
Ben-Josef
E
,
Bloomston
PM
,
Botha
JF
,
Clary
BM
, et al
NCCN clinical practice guidelines in oncology: hepatobiliary cancers
.
J Natl Compr Canc Netw
2009
;
7
:
350
91
.
31.
Moon
WS
,
Rhyu
KH
,
Kang
MJ
,
Lee
DG
,
Yu
HC
,
Yeum
JH
, et al
Overexpression of VEGF and angiopoietin 2: a key to high vascularity of hepatocellular carcinoma?
Mod Pathol
2003
;
16
:
552
7
.
32.
Schoenleber
SJ
,
Kurtz
DM
,
Talwalkar
JA
,
Roberts
LR
,
Gores
GJ
. 
Prognostic role of vascular endothelial growth factor in hepatocellular carcinoma: systematic review and meta-analysis
.
Br J Cancer
2009
;
100
:
1385
92
.
33.
Kaseb
AO
,
Hanbali
A
,
Cotant
M
,
Hassan
MM
,
Wollner
I
,
Philip
PA
. 
Vascular endothelial growth factor in the management of hepatocellular carcinoma: a review of literature
.
Cancer
2009
;
115
:
4895
906
.
34.
Kaseb
AO
,
Morris
JS
,
Hassan
MM
,
Siddiqui
AM
,
Lin
E
,
Xiao
L
, et al
Clinical and prognostic implications of plasma insulin-like growth factor-1 and vascular endothelial growth factor in patients with hepatocellular carcinoma
.
J Clin Oncol
2011
;
29
:
3892
9
.
35.
Pena
C
,
Lathia
C
,
Shan
M
,
Escudier
B
,
Bukowski
RM
. 
Biomarkers predicting outcome in patients with advanced renal cell carcinoma: results from sorafenib phase III treatment approaches in renal cancer global evaluation trial
.
Clin Cancer Res
2010
;
16
:
4853
63
.
36.
Bianchi
G
,
Loibl
S
,
Zamagni
C
,
Salvagni
S
,
Raab
G
,
Siena
S
, et al
Phase II multicenter, uncontrolled trial of sorafenib in patients with metastatic breast cancer
.
Anticancer Drugs
2009
;
20
:
616
24
.
37.
Zhu
AX
,
Dahani
DV
,
Duda
DG
,
di Tomaso
E
,
Ancukiewicz
M
,
Catalano
OA
, et al
Efficacy, safety, and potential biomarkers of sunitinib monotherapy in advanced hepatocellular carcinoma: a phase II study
.
J Clin Oncol
2009
;
27
:
3027
35
.
38.
Harmon
CS
,
DePrimo
SE
,
Raymond
E
,
Cheng
AL
,
Boucher
E
,
Douillard
JY
, et al
Mechanism-related circulating proteins as biomarkers for clinical outcomes in patients with unresectable hepatocellular carcinoma receiving sunitinib
.
J Transl Med
2011
;
9
:
120
.
39.
DePrimo
SE
,
Bello
CL
,
Smeraglia
J
,
Baum
CM
,
Spinella
D
,
Rini
BI
, et al
Circulating protein biomarkers of pharmacodynamic activity of sunitinib in patients with metastatic renal cell carcinoma: modulation of VEGF and VEGF-related proteins
.
J Transl Med
2007
;
5
:
32
.
40.
Willett
CG
,
Boucher
Y
,
Duda
DG
,
di Tomaso
E
,
Munn
LL
,
Tong
RT
, et al
Surrogate markers for antiangiogenic therapy and dose-limiting toxicities for bevacizumab with radiation and chemotherapy: continued experience of a phase I trial in rectal cancer patients
.
J Clin Oncol
2005
;
23
:
8136
9
.
41.
Drevs
J
,
Zirrgiebel
U
,
Schmidt-Gersbach
CIM
,
Mross
K
,
Medinger
M
,
Lee
L
, et al
Soluble markers for the assessment of biological activity with PTK787/ZK 222584 (PTK/ZK), a vascular endothelial growth factor receptor (VEGFR) tyrosine kinase inhibitor in patients with advanced colorectal cancer from two phase I trials
.
Ann Oncol
2005
;
16
:
558
65
.
42.
Baar
J
,
Silverman
S
,
Lyons
J
,
Fu
P
,
Abdul-Karim
F
,
Ziats
N
, et al
A vasculature-targeting regimen of preoperative docetaxel with or without bevacizumab for locally advanced breast cancer: impact on angiogenic markers
.
Clin Cancer Res
2009
;
15
:
3583
90
.
43.
Siegel
AB
,
Cohen
EI
,
Ocean
A
,
Lehrer
D
,
Goldenberg
A
,
Knox
JJ
, et al
Phase II trial evaluating the clinical and biologic effects of bevacizumab in unresectable hepatocelllular carcinoma
.
J Clin Oncol
2008
;
36
:
2992
8
.
44.
Loupakis
F
,
Cremolini
C
,
Fioravanti
A
,
Orlandi
P
,
Salvatore
L
,
Masi
G
, et al
Pharmacodynamic and pharmacogenetic angiogenesis-related markers of first-line FOLFOXIRI plus bevacizumab schedule in metastatic colorectal cancer
.
Br J Cancer
2011
;
104
:
1262
9
.
45.
Ebos
JM
,
Lee
CR
,
Christensen
JG
,
Mutsaers
AJ
,
Kerbel
RS
. 
Multiple circulating proangiogenic factors induced by sunitinib malate are tumor-independent and correlate with antitumor efficacy
.
Proc Natl Acad Sci USA
2007
;
104
:
17069
74
.
46.
Huang
H
,
Bhat
A
,
Woodnutt
G
,
Lappe
R
. 
Targeting the ANGPT-TIE2 pathway in malignancy
.
Nat Cancer Rev
2010
;
10
:
575
85
.
47.
Imanishi
Y
,
Hu
B
,
Jarzynka
MJ
,
Guo
P
,
Elishaev
E
,
Bar-Joseph
I
, et al
Angiopoietin-2 stimulates breast cancer metastasis through the α5β1 integrin-mediated pathway
.
Cancer Res
2007
;
67
:
4254
63
.
48.
Park
JH
,
Park
KJ
,
Kim
YS
,
Sheen
SS
,
Lee
KS
,
Lee
HN
, et al
Serum angiopoietin-2 as a clinical marker for lung cancer
.
Chest
2007
;
132
:
200
6
.
49.
Tanaka
S
,
Mori
M
,
Sakamoto
Y
,
Makuuchi
M
,
Sugimachi
K
,
Wands
JR
. 
Biologic significance of angiopoietin-2 expression in human hepatocellular carcinoma
.
J Clin Invest
1999
;
103
:
341
5
.
50.
Herceptin (trastuzumab) prescribing information
.
South San Francisco, CA
:
Genentech USA, Inc
.
[updated 2010 Oct; cited 2011 Nov 28]. Available from:
http://www.gene.com/gene/products/information/pdf/herceptin-prescribing.pdf.
51.
Erbitux (cetuximab) prescribing information, a wholly-owned subsidiary of Eli Lilly and Company, and Bristol-Myers Squibb Company
.
[updated 2011 Jun; cited 2011 Nov 28]. Available from:
http://packageinserts.bms.com/pi/pi_erbitux.pdf.
52.
Vectibix (panitumumab) prescribing information
.
Thousand Oaks, CA
:
Amgen Inc
.
[updated 2011 Jun; cited 2011 Nov 28]. Available from:
http://pi.amgen.com/united_states/vectibix/vectibix_pi.pdf.
53.
Zelboraf (vemurafenib) prescribing information
.
South San Francisco, CA
:
Genentech USA, Inc
.
[updated 2011 Aug; cited 2011 Nov 28]. Available from
: http://www.gene.com/gene/products/information/zelboraf/pdf/pi.pdf.
54.
Xalkori (crizotinib) prescribing information
.
Philadelphia, PA
:
Pfizer Inc
.
[updated 2011 Oct; cited 2011 Nov 28]. Available from:
http://labeling.pfizer.com/showlabeling.aspx?id=676.
55.
Kaposi-Novak
P
,
Lee
JS
,
Gomez-Quiroz
L
,
Coulouarn
C
,
Factor
VM
,
Thorgeirsson
SS
. 
Met-regulated expression signature defines a subset of human hepatocellular carcinomas with poor prognosis and aggressive phenotype
.
J Clin Invest
2006
;
116
:
1582
95
.
56.
Liu
L
,
Cao
Y
,
Chen
C
,
Zhang
X
,
McNabola
A
,
Wilkie
D
, et al
. 
Sorafenib blocks the RAF/MEK/ERK pathway, inhibits tumor angiogenesis, and induces tumor cell apoptosis in hepatocellular carcinoma model PLC/PRF/5
.
Cancer Res
2006
;
66
:
11851
8
.
57.
Kasahara
K
,
Arao
T
,
Sakai
K
,
Matsumoto
K
,
Sakai
A
,
Kimura
H
, et al
Impact of serum hepatocyte growth factor on treatment response to epidermal growth factor receptor tyrosine kinase inhibitors in patients with non-small cell lung adenocarcinoma
.
Clin Cancer Res
2010
;
16
:
4616
24
.