Hepatocellular cancer (HCC) is a rapidly fatal malignancy that is often diagnosed at an advanced stage when it is not curable. Despite a rise in incidence, this cancer is not as common in the Western world as it is in Asia and Africa.
Roswell Park Cancer Institute,
Elm and Carlton Streets
Buffalo, NY 14263
State University at Buffalo
Hepatocellular cancer (HCC) is a rapidly fatal malignancy that is often diagnosed at an advanced stage when it is not curable. Despite a rise in incidence, this cancer is not as common in the Western world as it is in Asia and Africa. Clinical research in this disease is challenging as trial-eligible patients are few. This is, in part, because most patients have several comorbidities, particularly hepatic dysfunction due to the underlying etiologic risk factors; there are also differences in cancer biology related to ethnicity, hepatitis, and cirrhosis that make designing and interpreting trials particularly challenging. However, in the last few years, important advances have been made both in understanding the molecular signatures of HCC and in completion of several well-designed, multicenter, phase III trials that have led to improvements in management of HCC. In addition to improved outcomes for patients, completion of these landmark phase III trials have also defined the way trials should be performed in these patients. Here, we summarize the evidence for current standards of care in managing HCC in the adjuvant and metastatic setting. The limitations of present therapies and important unanswered questions are summarized.Hepatocellular cancer (HCC) is a lethal malignancy that is an increasing global health problem. With an estimated global incidence of 749,000 cases in 2008, it is the 5th most common cancer and the 3rd most fatal.1Most cases occur where hepatitis B viral infection (HBV) is endemic; hepatitis C virus (HCV), alcoholic cirrhosis, and nonalcoholic steatohepatitis (NASH) are the other major causes. The lethality of this cancer stems, in part, from advanced stage at diagnosis, poor liver reserve, high rate of recurrence after curative therapies, and resistance to chemotherapy.
Hepatocelluar carcinoma has been a rare cancer in the United States, but there appears to be a clear trend toward its increasing incidence, which may be related to the HCV epidemic and the growing incidence of metabolic syndrome and obesity.2Although newer effective therapies for HCV will soon be available, the American Association for the Study of Liver Diseases (AASLD) estimates that as many as 4 million persons in the US are already chronically infected with HCV, and approximately half of them are unaware of their status. These individuals may have secondary cirrhosis. Detection of early HCV is difficult, and many may ultimately progress to advanced liver disease and/or HCC in the very near future.
Hepatocellular cancer is a complex and heterogeneous tumor driven by angiogenesis and several genomic alterations, including aberrant activation of signaling cascades such as EGFR, Ras/Erk, PI3K/ mTor, and apoptotic signaling.3Inflammation and immune dysfunction are hallmark symptoms of HCC.4
Chemotherapy has unfortunately not shown improved outcomes in patients with HCC and, due to underlying organ dysfunction, safety has been a concern, making it a secondary option to targeted therapy, especially in the Western world. This article summarizes the current approach to management of HCC, highlights trials that are expected to impact patient care in the near future, and reviews continuing needs in the field.The American Hepato-Pancreato-Biliary Association/American Joint Commission on Cancer conducted a consensus conference on HCC in 2002. The commission concurred that no single staging system can be used to accurately stage all patients with HCC.5Outcomes for patients with HCC are greatly influenced by tumor location, extrahepatic spread, vascular involvement, liver function, patient performance status, and comorbidities, making it very difficult to rely on any one staging system to predict prognosis. Each individual treatment option has clear eligibility and exclusions that are adhered to using multidisciplinary team discussions to individualize patient care when possible.
Of the available staging systems, the one most commonly used by clinicians is the Barcelona Cancer of the Liver Clinic (BCLC) algorithm.6This algorithm divides patients into 4 subsets: early, intermediate, advanced, and terminal. Clinicians also use the Child-Pugh Score to classify patients as having class A, B, or C liver cirrhosis, which has been adopted in clinical trials of novel therapies.There are many available therapeutic options for patients with HCC. These are briefly summarized inTable 1. The most commonly used therapy options are detailed below.
The entire liver is at risk for developing cancer due to inflammation, and chronic viral infection and cirrhosis often limit treatment options. Liver transplant treats both the cancer and the liver dysfunction. Although the BCLC staging system is widely used by clinicians and does provide criteria for liver transplant, the guidelines adhered to by the United Network for Organ Sharing (UNOS) vary. The UNOS guidelines state that liver transplant is reserved for patients with compromised liver function and limited tumor burden that meet the Milan/Mazzaferro criteria for liver transplantation for HCC (solitary tumor ≤5 cm or ≤3 tumors ≤3 cm with no macrovascular invasion).7
The outcomes of 1556 patients who underwent liver transplant at 36 centers were evaluated. In this analysis, 5-year survival for the 444 patients who met Milan criteria was 73% (95% CI: 68-78) compared with 54% (95% CI: 50-57) for the 1112 patients who exceeded the criteria.
Single lesion <5 cm or 3 lesions ≤3 cm and no vascular invasion or extrahepatic spread
70% 5 years
Usually limited to solitary tumors ≤5 cm; can be performed in select patients with larger tumors
40% relapse-free survival and 70%-90% OS at 5 years
4 cm size, max 3 lesions
40% 5-year survival
Multifocal tumors not amenable to ablation or surgery. May also be used as a bridge to transplant. Bilirubin <2 mg/dL, no main portal vein occlusion, no ascites, and no contraindication to use of iodinated IV contrast
Advanced, inoperable HCC with vascular invasion and or extrahepatic spread
6.5 months in Asian patients (most of whom have HBV) and 10.7 months in Western patients (HCV and NASH patients)
Due to shortage of organs and varying degrees of comorbidities and extent of liver dysfunction, patients with end-stage liver disease who are referred for transplant are given a Model for End-Stage Liver Disease (MELD) score and followed at a frequency based on expected change in their clinical status.8The MELD score is a prospectively developed and validated cirrhosis severity scoring system that uses a patient’s laboratory values for serum bilirubin, serum creatinine, and the international normalized ratio to predict 3-month survival using online calculators in patients with cirrhosis. An increased MELD score (the highest being 40) is associated with increased severity of hepatic dysfunction and increased 3-month mortality risk.
Surgery or local ablation can be potentially curative options when disease is limited to the liver and the procedure is technically feasible and safe. In patients who are able to undergo resection, 5-year relapse-free survival is ≈40% and overall overall survival (OS) may be as high as 90%.9Outcomes have been shown to correlate with tumor size. Although most surgeons restrict resection to patients with tumors <5 cm and are concerned about peritoneal seeding in patients who present with a capsular rupture, resection is still an option in such cases in select patients, although long-term survival is lower.
Radiofrequency ablation (RFA) is sometimes employed when surgery is contraindicated due to comorbidity or inadequate liver reserve for major hepatectomy. Radiofrequency ablation is limited to use in tumors sized <4 cm; most clinicians would also not employ RFA when there are >2 to 3 lesions. Patients with more lesions may be referred for chemoembolization or radioembolization.
Several other ablative techniques are now available. Electroporation (Nanoknife), microwave ablation, cryoablation, and alcohol ablation are all options based on availability and operator experience. Selection of one technique over another is based on size, proximity to blood vessels, and treatment center experience, but as of yet has not been compared in head-to-head trials.
Although effective antiviral therapy can delay HCC occurrence, outcomes following surgery for patients with HBV are poor. In a 2-stage longitudinal study in 780 patients, viral load ≥104was correlated with poor outcomes.10In a randomized cohort of 163 patients, antiviral therapy with lamivudine, adefovir dipivoxil, or entecavir significantly decreased HCC recurrence (hazard ratio [HR] 0.48; 95% CI: 0.32-0.70) and HCC-related death (HR 0.26; 95% CI: 0.14 to 0.50) in multivariate Cox analyses. With the recent availability of newer potent antiviral therapies for chronic HCV infection, similar trials need to be conducted.
Given that angiogenesis plays a role in early carcinogenesis and the targeted antiangiogenic sorafenib has shown efficacy in advanced HCC, the role of adjuvant sorafenib following surgery or RFA was evaluated in the international STORM study. The study accrued 1114 patients, 62% of whom were Asian.11Patients were randomized to sorafenib 400 mg PO BID or placebo until disease progression (up to 4 years). Sorafenib was not tolerable at the intended dose (median dose achieved was 578 mg QD); placebo was tolerated for only 2 years. No difference was seen in relapse-free survival in either group.
Transcatheter Arterial Chemoembolization
The blood supply of hepatocellular tumors is predominantly arterial, which allows for transarterial delivery of chemotherapy and cytotoxic effect from occlusion of tumor blood flow. Lipiodol was the agent initially used for transcatheter arterial chemoembolization (TACE); more commonly used agents now are doxorubicin, cisplatin, or doxorubicin-eluting microspheres (beads), which are employed with similar efficacy to each other.
Transcatheter arterial chemoembolization is employed for large, multifocal HCC that is not amenable to surgery or ablation, or as a bridge to liver transplant. When performed in carefully selected patients (bilirubin <2 mg/dL, with no ascites or main portal vein occlusion), it is generally safe. Postembolization syndrome consisting of fever, pain, anorexia, and fatigue are the common treatment-related side effects. Liver decompensation is rare but can occur, and contrast nephropathy due to the use of contrast has also been noted.
Therasse et al13
Change in the sum of the longest dimension of entire lesion and extrahepatic lesions. Measurable lesions are ≥1 cm at baseline and any new lesions count as progression.
CR, PR, SD, and PD defined by sum of longest dimension and new lesion appearance
Lencioni et al14
Change in the sum of the arterial enhancing area (viable tumor) thus excluding the area of dense lipiodol deposition, which indicates a necrotic area in the liver, including lesions that were ≥1 cm at baseline. Extraabdominal target lesions are assessed using RECIST as arterial phase is not obtained during chest scans.
CR, PR, SD, and PD defined by sum of longest dimension of viable tumor in abdomen and appearance of new lesion
Bruix et al15
EASL defines target and overall responses: target lesions as all measurable arterially enhancing lesions with diameters ≥1 cm in the liver. Target responses do not take into consideration changes in nontarget lesions and the appearance of new lesions. By contrast, overall responses assess all target lesions, but also take into consideration nontarget lesions (including measurable arterially enhancing lesions with diameters <1 cm, nonmeasurable arterially enhancing lesions in the liver, and extrahepatic disease sites), as well as the appearance of new lesions (including arterially enhancing lesions in the liver, and extrahepatic disease sites).
Patients exhibiting CR and PR are considered responders; those exhibiting SD and PD are considered nonresponders.
CR, complete remission; EASL, European Association for the Study of the Liver criteria; mRECIST, modified RECIST; PD, progressive disease; PR, partial remission; RECIST, Response Evaluation Criteria in Solid Tumors; SD, stable disease.
It is often challenging to assess response to TACE. Most centers use traditional RECIST criteria. In a recent report, response criteria using European Association for the Study of the Liver (EASL) (50.6%) and modified RECIST (mRECIST) (51.6%) were compared with standard RECIST. Much greater response rates were seen with EASL and mRECIST compared with RECIST (16.5%). The concordance was strong between the mRECIST and EASL results (k = 0.9) but weak between mRECIST and RECIST (k = 0.3). The EASL and mRECIST responses significantly correlated with survival. Risk reductions of 52% and 50% were observed for EASL and mRECIST responders, respectively, compared with nonresponders, suggesting that these criteria should be used more commonly. The results also suggest that median time to assessing response should be ≈90 days or 3 months for optimal assessment.12The key differences between these 3 response criteria are shown inTable 2.
Hepatocellular cancer is not considered a very radiosensitive tumor, but stereotactic body radiation (SBRT) is increasingly being used in conjunction with other options or when other options are not feasible. Radioembolization is the selective delivery of millions of yttrium-90 labeled radioactive glass microspheres (TheraSpheres) that are 20 to 30 micrometers in diameter by infusing them directly into the hepatic artery supplying blood to the liver tumors. This modality is also employed in centers that have this option available; however, it has been approved on a compassionate use as a therapeutic exemption due to lack of effective options. Contraindications are similar to TACE, and utility is limited due to lack of data on its role given other available local options.
Sorafenib is an oral multikinase inhibitor that inhibits Raf serine/threonine kinases and receptor tyrosine kinases (vascular endothelial growth factor receptors 1, 2, 3 and platelet-derived growth factor-β, Flt-3, and c-kit) that are implicated in tumorigenesis and tumor progression.16,17However, it is not yet known if this is the only mechanism of action.18
In the landmark phase III SHARP trial, sorafenib given at 400 mg PO BID was compared with placebo. Overall survival was 10.7 months in the sorafenib group and 7.9 months in the placebo group (HR 0.69; 95% CI: 0.55-0.87;P<.001).19This survival benefit was seen in the absence of radiographic shrinkage, and time to symptomatic progression was not significantly different between the 2 groups (4.1 months vs 4.9 months, respectively,P= .77). Most of the toxicities seen were grade 1 and 2; however, the incidence of treatment-related side effects was 80% in the sorafenib group and 52% in the placebo group and >60% of patients required dose interruptions or dose delays as a result of side effects. Main treatment-related toxicities were fatigue, hand-foot skin reaction, diarrhea, hypertension, and gastrointestinal side effects.
Although the phase II trial that led to the SHARP study included some Child-Pugh class B patients,20the patients selected for the SHARP trial were Child-Pugh class A and were clearly defined using the BCLC clinical staging and treatment algorithm, which has become widely accepted and used by clinicians. There is a consensus among HCC trial experts that future trials should include patients similar to those with advanced disease that were included in the SHARP trial, defined as BCLC class C (portal invasion, N1, M1 disease, or ECOG 1-2).
The global investigation of therapeutic decisions in HCC and of its treatment with sorafenib trial (GIDEON) collected postapproval data (phase IV) to assess the safety and utilization of sorafenib in actual clinical practice throughout the world. Initial reports from baseline enrollment and toxicity data showed that clinicians are using sorafenib in Child- Pugh class B and even C patients, and toxicity does not appear to be much higher in that population, although survival in those patients is generally short. Final efficacy results from this trial are expected in 2014.
Given the highly vascular nature of HCC tumors, the benefit from sorafenib therapy was not surprising. However, no validated biological markers currently exist for appropriately selecting patients who will benefit from therapy with sorafenib. There are no biomarkers identifying escape pathways that should be targeted after tumors develop resistance to sorafenib or any other anti angiogenic agent. As such, biomarker selection is a major focus of HCC research.21The search for surrogate markers for response has focused on circulating biomarkers, tissue biomarkers, and imaging biomarkers; however, none have yet emerged as being reliably predictive of benefit from anti-angiogenic agents, despite the fact that they are all mechanistically based.21
Since the approval of sorafenib for HCC in 2007, targeted agents with similar mechanisms of action that vary in their potency and toxicity profile have been tested in advanced HCC. However, none has been proven to improve outcomes significantly or to be more tolerable, highlighting the need to look at newer sorafenib-based combinations that will be tolerable and to develop novel targets beyond angiogenesis to improve outcomes.
In a large, international phase III trial, sunitinib 37.5 mg PO QD was compared with sorafenib with the primary endpoint being OS. The trial accrued 1074 patients, but was terminated early due to futility and safety reasons. Overall survival was 7.9 months in the sunitinib arm and 10.2 in the sorafenib arm (HR: 1.3; one-sidedP= .9990; 2-sidedP= .0014) Interestingly, survival in Asian patients and HBV-infected patients was similar in both the sorafenib and sunitinib arms, but in patients with HCV infection, survival with sunitinib was significantly shorter (9.2 vs 17.6 months; HR: 1.52; one-sidedP= .9835). Toxicity occurred more frequently in the sunitinib arm, although discontinuation of therapy due to toxicity was similar in both arms (≈13%).22
Brivanib is an oral anti-angiogenic agent that also inhibits fibroblast growth factor receptors. It was the first drug of this class to be evaluated in HCC both in the first-line and second-line settings. In the firstline trial, 1150 patients were accrued internationally and randomized to either sorafenib 400 mg PO BID or brivanib 800 mg PO QD.23 The primary endpoint of OS noninferiority was not met, and although toxicity rates seemed similar, discontinuation due to side effects occurred in 43% of patients in the sorafenib arm and in 53% in the brivanib arm.
A randomized, multicenter, placebo-controlled trial enrolled 395 patients who were intolerant of sorafenib or had failed on sorafenib. Patients were randomized 2:1 to brivanib or placebo.24 In this trial, median OS was 9.4 months in the brivanib group and 8.2 months for the placebo group (HR: 0.89; 95.8% CI: 0.69-1.15;P= .3307). Time to tumor progression and overall responses were not primary endpoints, but favored brivanib. Toxicities were as expected with this class of drug, with fatigue, hypertension, anorexia, and hyponatremia being the most frequent (≥10%).
Sorafenib vs placebo
Improved OS and TTP, not time to symptomatic progression.
Similar magnitude of benefit in Asian and Western populations, but 4-month lower survival in Asian trial for both placebo and treatment arms.
Sunitinib vs sorafenib
Sunitinib not superior to sorafenib in HBV-infected patients, but inferior to sorafenib in HCV-infected patients. Greater toxicities occurred with sunitinib.
Brivanib vs sorafenib
Brivanib did not meet noninferiority endpoint in OS compared with sorafenib.
Brivanib vs placebo (second-line therapy)
Brivanib survival was not statistically significantly higher than placebo. TTP and response favored brivanib.
OS indicates overall survival; TTP, time to progression.
The key take-home points from the 5 large phase III trials of targeted agents in the advanced setting that have been completed in the last 8 years are summarized inTable 3.
Sorafenib + Doxorubicin
Promising median TTP of 6.4 months and OS of 13.7 months has been seen with sorafenib 400 mg PO BID + doxorubicin 60 mg/m2 IV q 21 days in a multicenter, randomized phase II trial (doxorubicin alone was the comparator arm). In response, a multicenter phase III trial of sorafenib + doxorubicin compared with sorafenib alone has been initiated. Accrual is ongoing.
Sorafenib and Transarterial Chemoembolization
The optimal sequence of combining locoregional and systemic therapy is unclear. The toxicities of each therapy and difficulty assessing response to antiangiogenic therapy combined with optimal response assessment to TACE make it difficult to interpret trial data. In some studies, the timing of sorafenib given weeks after TACE may not be ideal. In other studies, the frequency of repeating TACE without clear reporting of response, along with the exclusion of patients with macrovascular invasion due to their ineligibility for TACE (thereby excluding more advanced patients) and disagreement on the ideal population for combining therapies (local disease only vs local plus systemic) make it difficult to compare data. Individual centers have protocols based on single-center experience that will remain standard practice until ECOG 1208 and SPACE trial data are available to answer these questions.
Immunomodulatory Therapies (Nivolumab, MK 3475, Bavituximab)
Studies performed by our group have shown that profound immunosuppression exists in patients with HCC compared with healthy normal subjects, and targeted depletion of subsets of immunosuppressive cells can partially restore antitumor immunity, suggesting that this may be a valid target in HCC.4,25At least 2 anti-PD-1 antibodies (nivolumab and MK 3475) are being evaluated or will soon be evaluated in patients with HCC with careful characterization of the hepatitis status and biomarkers that are eagerly awaited. Phosphatidylserine (PS) is a phospholipid which, unlike other phospholipids that are randomly distributed in the plasma membrane of most cells, is actively maintained by an ATP-dependent. Phosphatidylserine externalizes during apoptosis and can trigger immunosuppressive signals that inhibit autoimmune antitumor responses. Bavituximab is a chimeric PS-targeting antibody that is well tolerated and is being evaluated in combination with sorafenib. It appears to have promising activity thus far.
Tivantinib (ARQ197) is an oral inhibitor of Met, which is overexpressed in 45% of HCC cases and appears to be a valid therapeutic target. In a phase II trial of 106 previously treated patients, tivantinib showed promising activity in patients with high c-Met identified by immunohistochemistry, which has led to the initiation of a randomized phase III trial of second-line therapy with either tivantinib or placebo (2:1) in 303 HCC with high MET patients. Tivantinib will be administered at 120 mg PO BID.26
Until 2007, doxorubicin was the only chemotherapy agent that had achieved modest activity in HCC. Since the approval of sorafenib, it is not used often. However, given the poor outcomes of Asian patients with sorafenib and the lack of options for those who progress on or are intolerant to sorafenib, several investigators have looked at using systemic chemotherapy in its place. Of the regimens studied in prospective trials, FOLFOX (leucovorin calcium, fluorouracil, oxaliplatin) appears to be a reasonably welltolerated option and is commonly used.27
In summary, although only 2 of the 5 phase III trials recently completed in patients with advanced HCC have been successful, there are a number of promising trials that are ongoing or are soon to begin accrual. These trials have the potential to change the way patients with HCC are treated in the near future.
Presently, there are >15 sorafenib combination trials. Although there is concern that a number of oral tyrosine kinase inhibitor plus chemotherapy trials have been negative or poorly tolerated, the doxorubicin + sorafenib combination appears to be tolerable and to hold promise for improving outcomes beyond sorafenib alone based on the phase II trial data.
The tivantinib trial is a biomarker-driven trial; as such, it holds the potential of yielding data to enable selection of patients for therapy who have a high likelihood of clinically meaningful benefit. The success of that trial would require a successful companion diagnostic biomarker assay and could dramatically move forward the field of biomarker-driven research in HCC.
There are several anti-PD-1 antibodies in development as well. In studies, nivolumab appears to be further along than others. Nivolumab trials also evaluate the role of tissue-based biomarkers and, given the need for tissue for enrollment in such studies, consideration should be made for pretreatment biopsy even in patients who have biochemically or radiographically diagnosed HCC, especially in those who are trial candidates.
Sorafenib + MK-3475 will be evaluated in a Cancer Therapy Evaluation Program-supported phase I trial where safety and important response data will be obtained that may elucidate the mechanism of response in antiangiogenic + immunotherapy. This may require a redefinition of the optimal timing for assessing response, as well as response criteria themselves, as both classes of agents possess unique features of tumor response. Specifically, sorafenib therapy is associated with reduced vascularity and increase in necrotic tumor volume, while tumor enlargement due to infiltrating lymphocytes before seeing shrinkage is a feature of anti-PD-L1 therapy.
Because immune dysfunction from advanced cancer and hepatitis are hallmarks of HCC,4newer agents that target immune suppression are currently a major focus of HCC research;
More than 15 trials of sorafenib combinations versus sorafenib alone are ongoing, and several second-line trials of novel antiangiogenics following sorafenib failure are also ongoing. These trials are expected to yield results over the coming 3 to 5 years. Based on available phase II data, sorafenib + doxorubicin and tivantinib in MET-high HCC appear to have great promise.
The novel mechanism of action and safety of bavituximab thus far make this another promising agent in the field of immunotherapy. Studies by our group have shown that profound immune dysfunction exists in HCC and that targeted depletion of immunosupressive cells can restore anti tumor immunity, providing strong support that immunotherapies may have a large role in HCC.