The most favorable outlook for patients with hepatocellular carcinoma (HCC) is for those who can undergo liver transplantation or surgical resection, but they constitute a minority of patients. Novel interventional techniques and the use of the new targeted agent sorafenib (Nexavar) are improving the prognosis for the other patients with HCC, said representatives of the National Comprehensive Cancer Network (NCCN) Hepatobiliary Cancers Panel.
Panel Chair Al B. Benson III, MD, described the problem, noting that HCC is the fifth most common cancer worldwide. The incidence of HCC has more than doubled over the past 3 decades, which is largely attributable to the increased incidence of liver cirrhosis from hepatitis C viral (HCV) infection.
HCV infection predominates as a risk factor for HCC, accounting for nearly half the cases, Dr. Benson claimed, whereas hepatitis B virus (HBV) accounts for about 15%. Estimates are that 5% of patients are infected with both viruses and that 33% of patients with HCC have no viral infection (Di Bisceglie AM et al. Am J Gastroenterol 2003;98:2060–2063).
There are challenges associated with diagnosing HCC, according to Dr. Benson, Professor of Medicine at the Robert H. Lurie Comprehensive Cancer Center of Northwestern University, Chicago, Illinois. The role of a-fetoprotein (AFP) testing is included in the NCCN guidelines, although it has shortcomings as a diagnostic method, he stated. It does not correlate with size, stage, or prognosis, and its sensitivity and specificity vary according to the population.
Whereas AFP often is normal in patients with alcoholic liver disease and HCC, it is often elevated in chronic HBV infection without HCC. As liver-imaging studies to diagnose HCC, ultrasonography and computed tomography (CT) “are less than perfect,” particularly in the presence of cirrhosis, said Dr. Benson, although the sensitivity of CT can be improved by triphasic CT scanning. Magnetic resonance imaging (MRI) can play a role in the diagnosis of HCC, although positron emission tomography (PET) scanning is rarely useful in this setting.
“For diagnosis, look for the right clinical setting—cirrhosis or HBV without cirrhosis—and radiographic findings consistent with HCC,” he said. New or growing lesions should be screened, and there should be no evidence of an extrahepatic primary tumor (ie, a colonoscopy might be warranted).
The best treatment options for individual patients depend on prognostic staging systems, which evaluate the stage of the tumor at presentation, the degree of liver dysfunction, the patient’s performance status, and the potential for successful treatment. These prognostic indicators were described by Elin R. Sigurdson, MD, PhD, Head of Surgical Research, Attending Surgeon, and Senior Member at Fox Chase Cancer Center, Philadelphia, Pennsylvania.
There are multiple staging systems for localized disease,
and some overlap. Some of these staging systems include traditional tumor
indicators as well as comorbidities of cirrhosis, as reflected in measures of
liver function. The Child-Pugh classification (A, B, C) has been used for over
30 years to evaluate patients for comorbidities (encephalopathy; ascites; and
elevated albumin, prothrombin, and bilirubin levels) that affect survival. The
Child-Pugh classification is particularly useful for determining prognosis in
cirrhotic patients, Dr. Sigurdson pointed out.
“This is not a static classification. You can spend a few weeks getting a patient to a new classification if you treat the encephalopathy, ascites, and so forth. You can make most cirrhotic patients better for a while. Especially if a patient requires another type of surgical intervention (eg, cholecystectomy), it is worth the time spent to improve his/her functional status to reduce operative morbidity and mortality,” Dr. Sigurdson emphasized.
The Okuda staging system contains many of the same parameters that are used in the Child-Pugh classification system, as well as level of disease (ie, tumor size). The Italian program scoring system (CLIP) is built on the Child-Pugh classification system, with additional information about nodular status, a-fetoprotein level, and presence or absence of portal vein thrombosis (which portends a worse prognosis).
One of the most accepted staging systems today is the Barcelona Clinic Liver Cancer (BCLC) staging system (Table). It is endorsed by the American Association for the Study of Liver Diseases and has the best prognostic power for estimating survival with treatment. “Patients eligible for surgery or ablation need to be classified as BCLC stage A; those individuals classified as BCLC stage B can receive treatment with embolization; patients with BCLC stage C or stage D classification are eligible for clinical trials only,” Dr. Sigurdson noted.
Surgical resection has been the mainstay of treatment for HCC for many years, but only 15%–25% of all HCC patients will be eligible for some form of surgical intervention. Approximately 90% of HCC is diagnosed in cirrhotic patients in the US, Dr. Sigurdson noted, “and that is the challenge for surgery.” Five-year survival rates are consistently around 50% after liver resection; however, new hepatomas commonly develop in the remnant liver, largely due to the “field effect” of viral hepatitis infections. Operative mortality is less than 3% in properly selected patients, Dr. Sigurdson said.
Randomized controlled trials have shown that it is possible to occlude portal inflow without detriment to outcomes. High blood loss and portal vein involvement are associated with poorer outcomes. In addition, better 5-year survival was associated with segmental or anatomic resections, as opposed to resections in which tumor was enucleated or wedged out, she stated. This finding has some bearing on the use of radiofrequency ablation (RFA), which is essentially an enucleation technique, she said.
Orthotopic liver transplantation is the preferred treatment of small HCC lesions in cirrhotic patients. “It treats the tumor, the cirrhosis, and the risk of new lesions in the liver remnant after resection,” she pointed out. “Unfortunately, it is feasible in fewer than 5% of patients.”
Five-year survival in most transplant series approaches 75%, “which is better than any surgical series,” she noted. “The problem is the lack of sufficient numbers of donor organs and the comorbidities that develop while patients are waiting. You also have to deal with the viral load and its impact on the donor liver, and a high rate of recidivism among alcoholic patients following transplantation.”
Standard indications for transplantation according to the Milan criteria (highly effective predictors of which patients will die early because of underlying liver disease) include a lesion < 5 cm, two to three lesions < 3 cm, and severe underlying cirrhosis. These indications have been expanded at certain centers to include larger lesions < 6.5 cm, two to three nodules < 4.5 cm, or uncomplicated cirrhosis with or without RFA to control local disease.
Advances in interventional radiology have resulted in a number of regional catheter-based approaches—infusional therapy, chemoembolization, bland embolization, radioembolization (selective internal radiation therapy), and drug-eluting microspheres—and local ablative techniques (RFA, cryoablation, microwave ablation, interstitial laser therapy, ethanol injection, and acetic acid injection) to treat patients with HCC.
Chemoembolization restricts blood supply to the tumor, thus inducing tumor ischemia. It allows for prolonged dwell time of chemotherapeutic agents (10–25 times greater than with intravenous infusion) and decreased systemic toxicity. The approach, however, is still poorly standardized.
Chemoembolization has been shown to produce high response rates (60%–80%), but interpretation of response can be a challenge, Dr. Benson stated. In pooled studies to date, median 1-year survival was 72%, 2-year survival was 53%, and 3-year survival was 40%. “Survival can be good, but it decreases over time,” Dr. Benson noted.
The value of chemoembolization was shown in a randomized controlled trial of arterial embolization or chemoembolization versus symptomatic treatment in patients with unresectable HCC from the BCLC (Llovet JM et al. Lancet 2002;359:1734–1739). The study included 112 patients with unresectable HCC not suitable for curative treatment, with Child-Pugh class A or B and Okuda stage I or II disease. Patients were treated with regularly repeated arterial embolization (gelatin sponge), chemoembolization (gelatin sponge plus doxorubicin), or conservatively.
The best survival was observed with chemoembolization: 82% at 1 year and 63% at 2 years, compared with 75% and 50%, respectively, for arterial embolization (without chemotherapy) and 63% and 27%, respectively, for best supportive care.
Radioembolization is a newer, highly effective catheter-based technique that is being adopted by some specialists but is not yet universally available. NCCN has recently integrated radioembolization into the HCC guidelines, Dr. Benson reported. Radioembolization relies on the internal delivery of high-dose radiation to tumors by trapping microspheres in the capillary bed. The yttrium-90 microspheres emit b irradiation and are supplied in two formulations: TheraSphere® and SIR-Spheres. This method allows for limited penetration of b rays and spares the normal liver parenchyma.
Radioembolization differs from chemoembolization in that it does not rely on ischemia for effectiveness. There is, in fact, increasing evidence that ischemia may not be desirable as a method for treating tumors regionally, because it may upregulate several molecular factors that prevent tumor-cell apoptosis, Dr. Benson pointed out.
“We all agree that well-performed comparative studies of chemoembolization and radioembolization are needed. Chemoembolization has the longest ‘track record,’ and data suggest it offers a survival advantage for patients with HCC,” he said. “Data on the role of radioembolization in the treatment of HCC are rapidly accumulating, and it has advantages: It is not based on the induction of ischemia, and it generally can be performed as an outpatient procedure. Bland embolization (embolization without chemotherapy) is the least-favored embolization method because it is not clear whether its benefits are comparable to those of other modalities, Dr. Benson said.
There is currently great interest in local destruction of tumor using chemical (alcohol, acetic acid) ablation or thermal (microwave, radiofrequency) ablation, which is a good alternative to surgery for small-volume disease in the cirrhotic liver. The procedure can be performed by laparotomy, laparoscopy, or interventional radiology. Hospitalization time is shorter than with other techniques, and there is less damage to the cirrhotic liver, Dr. Sigurdson added.
Several randomized controlled trials have compared RFA with alcohol ablation (Lin SM et al. Gastroenterology 2004;127:1714–1723; Shiina S et al. Gastroenterology 2005;129:122–130; Lin CS. Jpn J Clin Oncol 2006;36:212–217). In all cases, RFA appeared to produce better relapse-free and overall survival and was accomplished in one or two treatments, versus four or five for alcohol ablation.
Pooled data show that RFA is superior to alcohol ablation in terms of local recurrence (13% vs 22%), distant recurrence (29% vs 59%), and 3-year survival (68% vs 60%). With either method, tumors smaller than 3 cm respond best, with complete destruction in 90%–95% of cases; tumors that are 3–5 cm are destroyed 60%–70% of the time, whereas those larger than 5 cm are poorly treated.
External-beam radiotherapy (3D conformal or stereotactic) allows focal treatment but is not considered curative. The technique is under investigation for HCC patients who are not candidates for resection or local destruction, including those with portal vein tumor thrombosis.
For some patients who are not transplant candidates, are inoperable, or have metastatic disease, more effective systemic treatment is now possible with the availability of sorafenib, Dr. Benson said.
He explained that HCC is extremely complex at the molecular
level. Chronic HBV/HCV infection or toxin exposure leads to cirrhosis and
predisposes hepatocytes to genetic damage. Increased cell proliferation and
survival result from abnormal growth factor stimulation (transforming growth
factor-beta, epidermal growth factor receptor), constitutive mitogenic
signaling (Raf/MEK/ERK, PI3K/Akt, Wnt), and deregulated antiapoptotic signaling
(p53, phosphatase and tensin homologue). Tumor cell stromal cell interactions
via soluble factors such as vascular endothelial growth factor (VEGF) promote
angiogenesis, which is required for tumor growth and progression. These
molecular abnormalities and others identified by gene-expression profiling are
candidate targets for inhibiting HCC initiation and progression, he said.
A better understanding of the multiple cellular signaling pathways implicated in the pathogenesis of HCC will reveal multiple potential therapeutic targets, Dr. Benson emphasized. However, he also cautioned that the complex interplay of these pathways could potentially provide a means of overriding the effects of an agent targeted to a specific molecular pathway.
Sorafenib is the agent that so far has emerged as important in HCC treatment, based on its preclinical activity and subsequent clinical trial results. The international phase III SHARP trial (Sorafenib HCC Assessment Randomized Protocol; Llovet JM et al. J Clin Oncol 2007; 25[18S]:LBA1) of 602 chemotherapy-naive patients with advanced HCC who were predominantly Child-Pugh class A was terminated early based on a 44% improvement in overall survival with sorafenib. The superiority of sorafenib was associated with a median survival of 46.3 weeks versus 34.3 weeks for placebo (P = 0.00058) and a longer time to disease progression, with a median of 24.0 weeks versus 12.3 weeks for placebo—a 73% prolongation (P = 0.000007). Interestingly, response rates were not much higher with sorafenib, but the disease progression-free rate was higher with sorafenib. The most pronounced benefit was seen in patients with HCV, who lived a median of 14 months with sorafenib treatment versus 7.9 months with placebo. Sorafenib was tolerable, with limited grade 3 and 4 toxicity, mainly diarrhea and hand-foot skin reactions.
Sorafenib is included in the NCCN guidelines as an option for Child-Pugh class A or B patients with inoperable HCC, but caution (ie, closer monitoring, etc.) should be used when administering sorafenib to Child-Pugh class B patients due to limited safety data in this population, Dr. Benson stated.
“Sorafenib is an important first step in using systemic therapy in these patients. We need more data with sorafenib, but it has opened the door,” Dr. Benson commented. Trials are now being designed that will evaluate sorafenib in combination with other biologics and with chemoembolization. Investigators are also evaluating the use of chemotherapy in the neoadjuvant setting for patients with HCC.
“Local regional therapies, as well as systemic therapy with sorafenib, have been most successful in patients classified with earlier Child-Pugh (eg, class A) and BCLC (eg, stage A) stages of disease,” he concluded.