109-132

Novel Therapeutics in Hepatocellular Carcinoma: How Can We Make Progress?

Author(s): 
Robin K. Kelley, MD, and
Alan P. Venook, MD
Article Summary: 


American Society of Clinical Oncology Educational Book

2013 ASCO Annual Meeting

Gastrointestinal (Noncolorectal) Cancer

The Many Faces of Hepatocellular Carcinoma: From Biology to Treatment

Overview

Hepatocellular carcinoma (HCC) is the third leading cause of cancer death globally, and its prevalence and impact are even more profound because sorafenib is the only systemic therapy proven to prolong survival in patients with advanced disease. Randomized phase III trials of other novel targeted agents including sunitinib, linifanib, brivanib, and the combination of sorafenib plus erlotinib have failed to improve overall survival compared with sorafenib as a single agent in the first line setting, as well as compared with placebo in the second-line setting, in the case of brivanib. These negative studies are a sobering reminder of the challenges to clinical research in HCC, including the competing comorbidity of liver dysfunction, marked clinical and biologic heterogeneity, and the unreliability of surrogate endpoints to accurately predict survival. To address these challenges, HCC-specific phase I/Ib cohorts must be used to define the maximum tolerated dose and drug exposure in this organ dysfunction population with high background rates of adverse events and little tolerance for superimposed treatment-related toxicity. Pooled analyses of contemporary randomized trials and database studies should be undertaken to define the strongest prognostic factors for stratification in future phase III studies. Research blood and archival tumor specimens should be collected from patients on clinical trials to intensify the search for biomarkers of responsive or resistant subsets, in parallel with ongoing efforts to improve on radiographic response assessment. Collectively, these and other new strategies are needed to make progress in identifying active novel therapeutics for patients with HCC.

Despite its enormous global impact as the third leading cause of cancer death worldwide, treatment options for advanced HCC remain extremely limited and outcomes grim.1 The multikinase inhibitor sorafenib is the only agent known to confer a survival benefit in advanced HCC, modestly prolonging overall survival compared with placebo (10.7 vs. 7.9 months, p < 0.001) in the SHARP trial.2 The failure of sunitinib, linifanib, brivanib, and the combination of sorafenib plus erlotinib to improve on sorafenib alone, and of brivanib to improve on placebo after first-line therapy, in recent comparative trials is a sobering reminder of the complexities of treating patients with HCC.3-7

The inter-related challenges in HCC of competing comorbidity and altered drug metabolism leading to high rates of treatment-related toxicity from underlying liver disease, clinical and biologic heterogeneity, and the unreliability of surrogate end points make the transition from phase II to phase III trials uncertain at best. Following on the reporting of these multiple recent negative studies, it is time to consider different strategies to make progress in identifying active novel therapeutics.

ACCOUNTING FOR THE COMPETING COMORBIDITY OF LIVER DISEASE IN CLINICAL TRIALS

A fundamental challenge in HCC is the coexistence of underlying liver injury and hepatic dysfunction as the premalignant environment in greater than 80% of patients.8,9 This pervasive problem is a competing cause of death but also has the potential to substantially affect drug metabolism and toxicity, as well as to increase the risk for a wide array of unrelated adverse events.

It is often difficult to discern the relationship of treatment to toxicity in HCC studies in the setting of hepatic dysfunction. Serious adverse event (SAE) rates (all-cause) reported in patients treated with single-agent sorafenib range from 36% to 55% in recent phase III trials,2,4,6,10 although this is not dissimilar to the rates of SAE reported in patients on placebo arms in these studies (45% to 57%).2,5,10 Overall, however, treatment-related event rates other than SAE generally appear higher in the experimental arms than the placebo arms across studies, though they have not been formally compared (Table 1). In recent sorafenib-controlled studies, toxicity in the experimental arms appears to parallel the experience in the placebo-controlled studies, with higher rates of toxicity and/or treatment discontinuation in experimental arms observed in relation to controls (Table 1).3,4,6,7

TABLE 1. Adverse Event Rates in Randomized Phase III Trials in HCC

RP3 Study and Treatment Arms All-Cause SAE (p) AE Differences Between Treatment Arms (p)
SOR versus placebo (SHARP)2 52% versus 54% (NS) 45% versus 32% (p = 0.04) (all-cause, treatment-emergent)
SOR versus placebo (Asia-Pacific)10 47.7% versus 45.3% (NS) 81.9% versus 38.7% (NR) (treatment-related, all grades)
SUT versus SOR4 44% versus 36% (NR) 82% versus 73% (NR) (all-cause, ≥ grade 3)
LIN versus SOR3 52.4% versus 38.5% (p < 0.001) 85.3% versus 75.0% (p < 0.001) (all-cause, ≥ grade 3)
SOR+ERLOT versus SOR6* 58% versus 54.6% (NS) 95.0% versus 95.2% (NS) (treatment-related, all grades)
BRIV versus SOR (BRISK-FL)7* 59% versus 52% (NR) NR
BRIV versus placebo (BRISK-PS)5 63% versus 57% (NR) 92% versus 62% (NR) (treatment-related, all grades)

Abbreviations: HCC, hepatocellular carcinoma; RP3, randomized phase III; SAE, serious adverse event; AE, adverse event; NS, not significant; NR, not reported; SOR, sorafenib; SUT, sunitinib; LIN, linifanib; ERLOT, erlotinib; BRIV, brivanib.

* Discontinuation rate due to adverse events was numerically higher on experimental arm (see Table 2).


Note: p values are provided only when reported.

Adverse events (AE) in HCC trials result in frequent dose-reductions, delays, and discontinuations across studies, which collectively suggest the potential to affect efficacy outcomes and mask identification of active therapies (Table 2). In the SHARP trial experimental and placebo arms, respectively, AE resulted in dose reductions in 26% and 7% and drug discontinuation in 38% and 37% (with 11% and 5% attributed as treatment-related events in each arm).2 The GIDEON registry (a global survey of the real-world use of sorafenib) findings corroborate that over one-third of patients with Child Pugh A (37%) require dose reductions.11 Similarly, sorafenib-controlled studies show that dose modifications also are required in a substantial proportion of all patients, perhaps more so in the experimental arms (Table 2).3,4,6,7 These trends further support the hypothesis that treatment-related toxicity, superimposed on high rates of background adverse events in liver dysfunction patients, may substantially affect dose intensity and exposure to novel therapeutics in clinical trials and confound efficacy analyses.

TABLE 2. Dose Reductions, Delays, and Discontinuations Due to AE in Randomized Phase III Trials in HCC

RP3 Study Dose Reductions (p) Discontinuations (p)
SOR versus placebo (SHARP)2 26% versus 7% (NR) 38% versus 37% (all-cause) (NR), 11% versus 5% (related) (NR)
SOR versus placebo (Asia-Pacific)10 30.9% versus 2.7% (NR) 19.5% versus 13.3% (all-cause) (NR)
SUT versus SOR4 47% versus 69% (NR) 26% versus 23% (all-cause) (NR)
LIN versus SOR3 45.3% versus 31.2% (p < 0.001) 36.3% versus 25.4% (all-cause) (p < 0.001)
SOR+ERLOT versus SOR6 NR 34.0% versus 23.8% (all-cause) (NR) (withdrawals after 1 cycle)
BRIV versus SOR (BRISK-FL)7 NR 43% versus 33% (all-cause) (NR)
BRIV versus placebo (BRISK-PS)5 NR 23% versus 7% (related) (NR)

Abbreviations: AE, adverse events; HCC, hepatocellular carcinoma; RP3, randomized phase III; NS, not significant; NR, not reported; SOR, sorafenib; SUT, sunitinib; LIN, linifanib; ERLOT, erlotinib; BRIV, brivanib.


Note: p values are provided only when reported.

TABLE 3. Response-Derived Endpoints and Overall Survival in Randomized Phase III Trials in HCC

RP3 Study DCR (p) TTP (p) OS (p)
SOR versus placebo (SHARP)2 43% versus 32% (p = 0.002) 5.5 versus 2.8 mo. (p < 0.001) 10.7 versus 7.9 mo. (p < 0.001)
SOR versus placebo (Asia-Pacific)10 35.3% versus 15.8% (p = 0.0019) 2.8 versus 1.4 mo. (p = 0.0005) 6.5 versus 4.2 mo. (p = 0.014)
SUT versus SOR4 NR 4.0 versus 4.1 mo. (NS) 8.1 versus 10.0 mo. (p = 0.0019)
LIN versus SOR3 13.0% versus 6.9% (NR) 5.4 versus 4.0 mo. (p = 0.001) 9.1 versus 9.8 mo. (NS)
SOR+ERLOT versus SOR6 43.9% versus 52.5% (NR) 3.2 versus 4.0 mo. (p = 0.91) 9.5 versus 8.5 mo. (p = 0.2)
BRIV versus SOR (BRISK-FL)7 12.0% versus 8.8% (NR) 4.1 versus 4.2 mo. (p = 0.85) 9.5 versus 9.9 mo. (NS)
BRIV versus placebo (BRISK-PS)5 79.2% versus 49.1% (p < 0.0001) 4.2 versus 2.7 mo. (p = 0.0001) 9.4 versus 8.2 mo. (p = 0.33)

Abbreviations: HCC, hepatocellular carcinoma; RP3, randomized phase III; NS, not significant; NR, not reported; DCR, disease control rate; TTP, time to progression; OS, overall survival; mo., months; SOR, sorafenib; SUT, sunitinib; LIN, linifanib; ERLOT, erlotinib; BRIV, brivanib.


Note: p values are provided only when reported.

Acknowledging liver disease as an immutable baseline factor in HCC populations, how can we reduce the degree of superimposed treatment-related toxicity in HCC clinical trials? In the trials cited in Tables 1 and 2, the high rates of all-cause and treatment-related toxicity and dose modifications occurred despite stringent eligibility criteria, which restricted enrollment to patients with Child-Pugh A liver disease, a clinical trial strategy to minimize background comorbidity rates in HCC clinical trials.12 Though improved tools to risk-stratify baseline liver disease could reduce the competing comorbidity of liver disease in patients entering clinical trials, such approaches, which further winnow the eligible HCC population, would adversely affect accrual rates as well as limit the applicability to the broader advanced HCC population.

Instead, these data suggest that the maximum tolerated dose (MTD) in all-comer phase I trials may exceed that in HCC populations. This hypothesis supports disease-specific dose-finding and pharmacokinetic (PK) and pharmacodynamics (PD) studies in HCC cohorts, instead of the conventional model of extrapolating from doses and PK/PD data in all-comer phase I studies. Conventionally, phase II and III efficacy studies in patients with HCC have been undertaken by using doses identified from all-comer phase I studies, without HCC-specific phase I studies. It is certainly possible that assessment of efficacy could be confounded by administration of doses exceeding the MTD in an HCC population.

Another important question pertaining to the study of investigational therapeutics in populations of patients with HCC is whether all-cause organ dysfunction studies accurately recapitulate the cumulative effect of cirrhosis on hepatic metabolism across drugs and metabolic pathways. The original studies of chemotherapeutics (paclitaxel and gemcitabine) in patients with organ dysfunction led the National Cancer Institute Organ Dysfunction Working Group (ODWG) to define liver dysfunction by aspartate aminotransferase and bilirubin levels in all-cause hepatic dysfunction populations.13-15 These criteria show modest association with Child Pugh score16 but have not been comprehensively examined for association with toxicity, PK, or PD in cirrhotics compared with patients with other causes of liver dysfunction (such as hepatitides, metastatic disease, biliary obstruction, or hepatotoxicity from previous chemotherapy). This important unanswered question suggests a role for future studies to determine concordance of ODWG criteria with the degree of hepatic dysfunction in HCC and cirrhotic populations in general.

CLINICAL HETEROGENEITY OF HCC: TO LUMP OR TO SPLIT, AND HOW?

Identifying active novel therapeutics in advanced HCC is also obscured by the marked heterogeneity in outcomes across populations and studies. This heterogeneity prohibits meaningful cross-study comparisons and mandates randomized trials to interpret efficacy end points. There is no more graphic example of the hazards of cross-study comparison in HCC than when comparing the SHARP trial2 with the contemporary Asia-Pacific study.10 Both were phase III trials of sorafenib compared with placebo in advanced HCC patients, but were conducted in mostly Western compared with Asia-Pacific populations, respectively. Despite the two studies' shared industry sponsorship, trial design, and eligibility, along with the achievement of nearly identical positive hazard ratios (HRs) for improvement in overall survival (HR 0.69 and 0.68, respectively), their absolute outcomes diverged substantially, with median overall survival for sorafenib compared with placebo arms of 10.7 compared with 7.9 months in SHARP by comparison to 6.5 compared with 4.2 months in the Asia-Pacific study.2,10 It remains unknown whether this disparity is a result of inherent differences in HCC biology dependent on etiology of liver disease (e.g., hepatitis B compared with C virus infection, alcoholism, nonalcoholic fatty liver disease), pharmacogenetic differences between Asian and Western populations, variations in regional practice patterns in early-stage disease and/or post-sorafenib, or a complex interplay of these and many other factors.

This finding exemplifies why stratification within randomized trials is particularly important to ensure that clinical and pathologic prognostic factors are balanced. In the phase III BRISK-PS trial of brivanib compared with placebo after failure of sorafenib therapy in advanced HCC after first-line therapy, an imbalance between experimental and placebo arms on vascular invasion (31% compared with 18% favoring placebo arm) was identified despite stratification on extrahepatic spread and/or vascular invasion, raising the question of whether this imbalance could have contributed to the failure of brivanib to demonstrate survival benefit in this study—something we will never know.5

Collectively, these examples underscore the potential effect of stratification factors on outcomes in HCC efficacy studies. From these contemporary randomized phase III datasets, retrospective pooled subset analyses as well as longitudinal database studies of putative and known clinical and pathologic prognostic factors should be undertaken to inform selection of the most impactful stratification factors in future HCC phase III trials. There remains also the important but unanswered question of whether future phase III trials of novel therapeutics in HCC should be conducted separately in Asian and Western populations, and if so, how should these populations be defined?

COMPLEX TUMOR BIOLOGY: A GOOD BIOMARKER IS HARD TO FIND

Complex tumor biology in HCC underlies the well-characterized clinical heterogeneity of HCC. HCC tumors demonstrate a high degree of genetic instability and heterogeneity attributed to longstanding inflammation and hepatocyte regeneration in the setting of chronic liver injury.17 Gene expression profiling in HCC suggests the existence of common molecular subclasses in HCC, including distinct groups with greater genetic instability, high grade histology, and poor clinical outcome.17-21 Recent comprehensive genotyping efforts in HCC tumors have identified a high rate of somatic mutation without frequently recurring activation of known driver oncogenes; however, a sobering common theme in carcinogen-induced cancers.22-25 An update on the complex mechanisms of hepatocellular carcinogenesis is presented separately in Dr. Philip Johnson's companion manuscript in this Educational Book.

In the setting of complex tumor biology, the paucity of available tumor tissue for the identification of biomarkers of response or resistance in HCC makes matters worse. The difficulty in obtaining research tumor specimens in HCC can be attributed, in part, to the acceptance of radiographic diagnosis without biopsy in HCC as well as to the scant material often obtained from fine needle aspirations when biopsies are performed, because of concerns for bleeding and biopsy track seeding.8 The SHARP trial again serves as an informative example, having successfully collected baseline and on-treatment research blood samples for biomarker research in 81.6% and 50.7% of the overall study population, respectively, by comparison with collection of archival tumor tissue samples in only 18%.26,27

In addition to ongoing efforts to identify circulating biomarkers in HCC, intensified efforts to collect clinically annotated HCC tumor specimens are essential to improving our understanding of HCC tumor biology and identifying biomarkers of response or resistance to novel therapeutics. These specimens should be obtained not only in the context of embedded research biopsies for clinical trials under the auspices of informed consent, but also as a standard of care if criteria for radiographic diagnosis are not stringently met.28

The selective MET inhibitor, tivantinib (ARQ 197) is an example of a novel therapeutic with potential activity identified by a tumor tissue biomarker enrichment strategy in HCC. In a randomized phase II study, patients with high MET expression by immunohistochemistry (IHC) treated with tivantinib (22 patients) compared with placebo (15 patients) had significantly longer time to progression (TTP) (2.7 vs. 1.4 months, p = 0.03).29 Based on this randomized phase II data, a phase III trial of tivantinib compared with placebo in second-line HCC patients with high MET expression is ongoing (NCT01755767). Another example of a promising biomarker in HCC is glypican-3 (GPC3), being examined in a randomized, placebo-controlled phase II trial (NCT01507168) as a predictive marker for a recombinant monoclonal antibody targeting GPC3, GC33, based on phase I data suggesting longer TTP in patients with GPC3-high disease.30 Table 4 presents a selection of new targeted agents under study in HCC. These examples in HCC highlight the potential of biomarker enrichment to identify active novel therapeutics and again underscore the importance of obtaining tumor tissue specimens from patients enrolled on therapeutic clinical trials.

TABLE 4. Selected New Agents and Targets in Clinical Development in HCC

Target Agent Trial Design* Trial
Arginine deprivation ADI-PEG20 (pegylated arginine deiminase) RP3 NCT01287585
GPC-3 GC33 (anti-glypican 3 Ab) RP2, stratified on GPC-3 IHC NCT01507168
Immune modulation JX-594 (oncolytic poxvirus)
    Lenalidomide (IMiD)
    Tremelimumab (anti-CTLA4 Ab)
RP2
Phase II
Phase II
NCT01387555
NCT01545804
NCT01008358
MET Cabozantinib (multikinase inhibitor)
    Tivantinib (MET inhibitor)
RP3
RP3, MET-high population
NCT pending
NCT01755767
mTOR CC-223 (TORC1/2 inhibitor)
    Everolimus (mTOR inhibitor)
Phase I/II
RP3
NCT01177397
NCT01035229
VEGFR Cabozantinib (multikinase inhibitor)
    Ramucirumab (anti-VEGFR Ab)
RP3
RP3
NCT pending
NCT01140347

Abbreviations: HHC, hepatocellular carcinoma; RP3, randomized phase III; RP2, randomized phase II; Ab, antibody; IHC, immunohistochemistry; VEGFR, vascular endothelial growth factor receptor.

Targets and agents are listed alphabetically. Key: *=All studies listed are for ≥ second-line therapy.

INCONSISTENT SURROGATE ENDPOINTS IN HCC TRIALS

The negative trials of linivanib, brivanib, sunitinib, and sorafenib plus erlotinib collectively draw attention to the definitions of go/no-go signals in phase II efficacy studies in HCC. To address the common criticism that these trials did not use randomization in phase II before proceeding to phase III, it is important to note that, even in the larger randomized phase III studies, conventional phase II response-based surrogate end points did not reliably predict overall survival and were sometimes discordant, even within the same study population (Table 3).3-7

Why are drugs failing to improve overall survival despite trends toward or even achieving an improvement in response-based end points? One possible explanation is simple: the drugs truly are not sufficiently active to improve survival and the surrogate end points failed as a result of limitations of standard imaging to discern response compared with progression in HCC, compounded by the inherent inability to adequately blind treatment arms when treating with drugs which confer characteristic symptomatic toxicities. The limitations of conventional response imaging in HCC are widely recognized, which is a discussion beyond the scope of this manuscript.12 Another possible explanation for the unreliability of surrogate end points in HCC hearkens back to the discussion above on the competing comorbidity of liver disease in HCC: perhaps the drugs or combinations do have activity as suggested by response-based end points, but dose modifications and discontinuations as a result of toxicity have contaminated the overall survival results. This concern reinforces the need for disease-specific phase I/Ib trials in HCC in an effort to minimize the effect of differences in drug exposure on efficacy outcomes. A third possibility is that the drugs or combinations may have activity only in subsets and that this activity is obscured by the heterogeneity of randomized phase III populations. This hypothesis supports the need to optimize stratification and identify biomarkers of responsive or resistant subpopulations as discussed above, as well.

Most likely, a complex combination of these factors is at play, which may vary according to individual drugs and clinical trials. Although a clear go/no-go signal for phase II trials in HCC remains difficult to define in the absence of reliable surrogate end points, an appropriate interim strategy is to use randomized phase II designs, careful stratification and biomarker efforts, and attention to overall survival as a secondary end point to inform the interpretation of phase II trial findings.

CONCLUSION

In summary, future progress in HCC clinical research requires a multifaceted approach to address the unique challenges in this disease including underlying liver dysfunction, the high degrees of clinical and biologic heterogeneity, and inconsistent surrogate end points for survival. HCC-specific phase I/Ib cohorts should be used instead of extrapolating from all-comer phase I data to define MTD, PK, and PD in this organ dysfunction population with high background rates of adverse events and little tolerance for superimposed treatment-related toxicity. Pooled analyses of contemporary randomized trials and database studies should be undertaken to define the strongest prognostic factors for stratification in future phase III studies. Research blood and archival tumor specimens must be collected from patients on clinical trials to intensify the search for biomarkers of responsive or resistant subsets, in parallel with ongoing efforts to improve on radiographic response assessment.

Key Points

  • Identification of active therapies in hepatocellular carcinoma (HCC) is challenged by the competing comorbidity of underlying liver disease, high degrees of clinical and biologic heterogeneity, and unreliable surrogate endpoints for overall survival.

  • The competing comorbidity of liver disease contributes to high rates of adverse events and dose modifications in HCC studies, and relationship to treatment can be difficult to discern. HCC-specific phase I dose-finding, pharmacokinetic, and pharmacodynamics studies should be undertaken before phase II and III efficacy studies whose endpoints may be confounded by a high incidence of toxicity and dose modifications.

  • The highly heterogeneous clinical outcomes across populations of patients with HCC prohibit meaningful cross-study comparisons and require randomization and stratification in efficacy studies. Pooled analyses and large, longitudinal database studies are needed to identify the optimal stratification factors for phase III trials.

  • Complex tumor biology requires intensified efforts to collect annotated HCC tumor specimens as well as to identify circulating biomarkers.

  • Surrogate endpoints have not reliably predicted overall survival in phase II as well as randomized phase III settings. Overall survival remains the gold standard for determining efficacy in the absence of reliable surrogate endpoints in HCC.

References

1. Parkin DM, Bray F, Ferlay J, at al. Global cancer statistics, 2002. CA Cancer J Clin. 2005;55:74-108.
PubMed | CrossRef
2. Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med. 2008;359:378-390.
PubMed | CrossRef
3. C Cainap QS, Huang W-T, Chung I-J, et al. Phase III trial of linifanib versus sorafenib in patients with advanced hepatocellular carcinoma (HCC). J Clin Oncol. 2013 Gastrointestinal Cancers Symposium; San Francisco, CA.
4. Cheng A YK, D Lin, J Park, et al. Phase III trial of sunitinib (Su) versus sorafenib (So) in advanced hepatocellular carcinoma (HCC). J Clin Oncol. 2011 American Society of Clinical Oncology Annual Meeting. Chicago, IL.
5. Llovet JM TD, J-L Raoul, E Boucher, et al. Brivanib versus placebo in patients with advanced hepatocellular carcinoma (HCC) who failed or were intolerant to sorafenib: results from the phase 3 BRISK-PS study. 2012 International Liver Congress, European Association for the Study of the Liver; Barcelona, Spain.
6. Zhu AX, Rosmorduc O., Evans J, et al. SEARCH: a phase III, randomized, double-blind, placebo-controlled trial of sorafenib plus erlotinib in patients with hepatocellular carcinoma (HCC). 2012 37th ESMO Congress; Vienna, Austria. European Society of Medical Oncology.
7. Johnson PJea. Brivanib (BRI) versus sorafenib (SOR) as first-line therapy in patients with unresectable, advanced hepatocellular carcinoma (HCC): results from the phase 3 BRISK-FL study. The 2012 Liver Meeting; Boston, MA.
8. Bruix J, Sherman M. Management of hepatocellular carcinoma: an update. Hepatology. 2011;53:1020-1022.
PubMed | CrossRef
9. Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma. Lancet. 2003;362:1907-1917.
PubMed | CrossRef
10. Cheng AL, Kang YK, Chen Z, 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.
PubMed | CrossRef
11. JA Marrero LR, Kudo M, et al. Global investigation of therapeutic decisions in hepatocellular carcinoma and of its treatment with sorafenib (GIDEON) second interim analysis in more than 1,500 patients: clinical findings in patients with liver dysfunction. J Clin Oncol. 2011 American Society of Clinical Oncology Annual Meeting; Chicago, IL.
12. Llovet JM, Di Bisceglie AM, Bruix J, et al. Design and endpoints of clinical trials in hepatocellular carcinoma. J Natl Cancer Inst. 2008;100:698-711.
PubMed | CrossRef
13. Ramanathan RK, Egorin MJ, Takimoto CH, et al. Phase I and pharmacokinetic study of imatinib mesylate in patients with advanced malignancies and varying degrees of liver dysfunction: a study by the National Cancer Institute Organ Dysfunction Working Group. J Clin Oncol. 2008;26:563-569.
PubMed | CrossRef
14. Venook AP, Egorin MJ, Rosner GL, et al. Phase I and pharmacokinetic trial of paclitaxel in patients with hepatic dysfunction: Cancer and Leukemia Group B 9264. J Clin Oncol. 1998;16:1811-1819.
PubMed
15. Venook AP, Egorin MJ, Rosner GL, et al. Phase I and pharmacokinetic trial of gemcitabine in patients with hepatic or renal dysfunction: Cancer and Leukemia Group B 9565. J Clin Oncol. 2000;18:2780-2787.
PubMed
16. Patel H, Egorin MJ, Remick SC, et al. Comparison of Child-Pugh (CP) criteria and NCI organ dysfunction working group (NCI-ODWG) criteria for hepatic dysfunction (HD): implications for chemotherapy dosing. J Clin Oncol. 2004 American Society of Clinical Oncology Annual Meeting.
17. Hoshida Y, Toffanin S, Lachenmayer A, et al. Molecular classification and novel targets in hepatocellular carcinoma: recent advancements. Semin Liver Dis. 2010;30:35-51.
PubMed | CrossRef
18. Katoh H, Ojima H, Kokubu A, et al. Genetically distinct and clinically relevant classification of hepatocellular carcinoma: putative therapeutic targets. Gastroenterology. 2007;133:1475-1486.
PubMed | CrossRef
19. Laurent-Puig P, Legoix P, Bluteau O, et al. Genetic alterations associated with hepatocellular carcinomas define distinct pathways of hepatocarcinogenesis. Gastroenterology. 2001;120:1763-1773.
PubMed | CrossRef
20. Lee JS, Chu IS, Heo J, et al. Classification and prediction of survival in hepatocellular carcinoma by gene expression profiling. Hepatology. 2004;40:667-676.
PubMed | CrossRef
21. Hoshida Y, Nijman SM, Kobayashi M, et al. Integrative transcriptome analysis reveals common molecular subclasses of human hepatocellular carcinoma. Cancer Res. 2009;69:7385-7392.
PubMed | CrossRef
22. Fujimoto A, Totoki Y, Abe T, et al. Whole-genome sequencing of liver cancers identifies etiological influences on mutation patterns and recurrent mutations in chromatin regulators. Nat Genet. 2012;44:760-764.
PubMed | CrossRef
23. Guichard C, Amaddeo G, Imbeaud S, et al. Integrated analysis of somatic mutations and focal copy-number changes identifies key genes and pathways in hepatocellular carcinoma. Nat Genet. 2012;44:694-698.
PubMed | CrossRef
24. Govindan R, Ding L, Griffith M, et al. Genomic landscape of non-small cell lung cancer in smokers and never-smokers. Cell. 2012;150:1121-1134.
PubMed | CrossRef
25. Greenman C, Stephens P, Smith R, et al. Patterns of somatic mutation in human cancer genomes. Nature. 2007;446:153-158.
PubMed | CrossRef
26. Llovet JM, Pena CE, Lathia CD, et al. Plasma biomarkers as predictors of outcome in patients with advanced hepatocellular carcinoma. Clinical Cancer Research. 2012;18:2290-2300.
PubMed | CrossRef
27. Llovet JM. Molecular signature of prognosis and outcomes in hepatocellular carcinoma. 2012 Gastrointestinal Cancers Symposium; San Francisco, CA.
28. Mullhaupt B, Durand F, Roskams T, et al. Is tumor biopsy necessary? Liver Transpl. 2011;17 Suppl 2:S14-25.
PubMed | CrossRef
29. Santoro A, Rimassa L, Borbath I, et al. Tivantinib for second-line treatment of advanced hepatocellular carcinoma: a randomised, placebo-controlled phase 2 study. Lancet Oncol. 2013;14:55-63.
PubMed | CrossRef
30. Zhu AX, Gold PJ, El-Khoueiry AB, et al. First-in-man phase I study of GC33, a novel recombinant humanized antibody against glypican-3, in patients with advanced hepatocellular carcinoma. Clin Cancer Res. 2013;19:920-928.
PubMed | CrossRef