Surveillance for hepatocellular carcinoma: in whom and how?

Epidemiology of HCC

Hepatocellular carcinoma (HCC) accounts for 85–90% of all primary liver cancer. Each year, approximately half a million people worldwide are newly diagnosed with HCC. A similar number die of HCC, thus demonstrating the dismal prognosis of this cancer. The highest HCC incidence rates are reported from regions endemic for hepatitis B virus (HBV) infection, including Southeast Asia (e.g. North and South Korea, China, and Vietnam) and sub-Saharan Africa (e.g. Mozambique). Chronic HBV and hepatitis C virus (HCV) explain the majority of HCC cases. In the United States, HCC is currently the fastest growing cause of cancer-related death. As the incidence rates for HCC have tripled in the past two decades [El-Serag and Rudolph, 2007], the distribution of HCC has shifted towards white Hispanic and non-Hispanic individuals of younger ages (45 and 60 years). This increase is at least partially attributable to the rise in HCV-related HCC [El Serag, 2002].

Survival is poor in most patients with HCC (5-year survival less than 5%) except in patients in the early stage who receive potentially curative therapy (liver transplant, surgical resection, or ablation), where a considerable improvement in survival has been observed (5-year survival ranges between 40% and 70%). However, these treatments are received by only a small number of patients. For example, population-based studies in the United States indicate that only approximately 10% of patients with HCC receive these treatments [El Serag et al. 2006]. Therefore, HCC surveillance has been advocated to detect HCC at an early stage, when critical treatment can be applied.

HCC surveillance recommendations

Who is targeted?

Major etiological risk factors for HCC include HBV, HCV, alcoholic liver disease, and, possibly, nonalcoholic fatty liver disease. These risk factors lead to the formation and progression of cirrhosis, which is present in about 80–90% of HCC patients. The 5-year cumulative risk of developing HCC in patients with cirrhosis ranges between 5% and 30% depending on etiology (highest in HCV), region or ethnicity (highest in Asians), and stage of cirrhosis (highest in decompensated disease) [Fattovich et al. 2004]. Several factors have been associated with increased HCC risk among patients with cirrhosis; although these factors have been tested as determinants for threshold of employing surveillance nor incorporated in practice guidelines, they may influence decision making for offering (physician) and accepting (patient) HCC surveillance. For example, among patients with HCV-related cirrhosis, the presence of older age, male sex, high BMI, diabetes, coinfection with HBV and HIV, active viremia or unsuccessful previous antiviral therapy, and heavy alcohol drinking are associated with increased HCC risk as compared with cirrhotic patients who do not have these factors.

Practice guidelines from the American Association of the Study of Liver Diseases and the European Association for the Study of the Liver have recommended HCC surveillance for patients at high risk of developing HCC [Zhang et al. 2004; Bruix et al. 2001]. Patients at high risk include those with cirrhosis and those with chronic HBV infection irrespective of cirrhosis (Table 1). HCC surveillance is not recommended in patients with HCV in the absence of cirrhosis.

Table 1.

High-risk groups for hepatocellular carcinoma (HCC) in whom surveillance might be indicated. Estimates of the annual HCC risk are also provided where reliable data are available.

Cirrhosis	HCC risk per year
Hepatitis C	2–7%
Hepatitis B	3–5%
Genetic hemochromatosis	NA
Primary biliary cirrhosis	2–3%
Non alcoholic steatohepatitis	NA
Alpha 1 antitrypsin deficiency, autoimmune hepatitis	NA
Hepatitis B carriers without cirrhosis (HBV surface Ag+)
Asian males >40 years of age	0.4–0.6
Asian females >50 years of age	0.3–0.6
Africans >20 years of age	NA
Family history of HCC	NA

How should HCC surveillance be performed?

We recommend a combination of liver ultrasound and serum alpha fetoprotein (AFP). Liver ultrasound is recommended as the primary surveillance modality for HCC; it has a modest sensitivity of approximately 60% and a higher specificity of approximately 85–90% [Singal et al. 2009]. The performance of ultrasound as a surveillance test depends on the experience of the examiner. It is reportedly less accurate in patients who are obese and those with a nodular liver [Tong et al. 2001]. Computed axial tomography (CT) and magnetic resonance imaging (MRI) scans have not been adequately tested for HCC surveillance, but are used in clinical practice. CT and MRI are associated with increased detection of more HCC than ultrasound but they are also associated with a higher false-positive rate [Kobayashi et al. 1985]. In addition, the high costs and potential harm related to radiation exposure and contrast-related injury associated with these tests make them poor candidates for surveillance tests in most practice settings.

AFP measurement is also commonly used for HCC surveillance because it is relatively inexpensive, simple to perform, and is widely available. However, AFP alone (without liver ultrasound) is not recommended as a HCC surveillance test due to its low sensitivity and specificity for detecting HCC. At a serum cutoff level of 20 ng/ml, AFP has low sensitivity ranging from 25% to 65% for detecting HCC [Paul et al. 2007]. Patients with chronic liver disease, especially those with a high degree of hepatocyte regeneration (e.g. HCV), can express elevated serum AFP in the absence of malignancy [Franca et al. 2004; Bayati et al. 1998]. Furthermore, only one third of patients with HCC have AFP levels higher than 100 ng/ml [Torzilli et al. 1999; Ebara et al. 1989].

Other tests that have been used and examined for HCC surveillance include des-gamma carboxy-prothrombin (DCP) and lectin-bound AFP (AFP-L3%). The evidence is inconclusive about whether these biomarkers perform better than AFP for detecting HCC. In a recent US study, AFP was found to be more sensitive than DCP and AFP-L3% for the diagnosis of early stage HCC at a new cutoff of 10.9 ng/ml [Marrero et al. 2009]. Other studies have found that DCP has a higher sensitivity for detecting and predicting the development of HCC [Durazo et al. 2008; Shimizu et al. 2002]. In a nested case–control study of patients who developed HCC during follow up of the HALT-C cohort, the sensitivity and specificity of AFP collected at the time of HCC diagnosis was 61% and 81% at a cutoff of 20 ng/ml and 22% and 100% at a cutoff of 200 ng/ml. Levels measured 1 year before the diagnosis of HCC, the sensitivity and specificity at the low cutoff was 43% and 94%, respectively, for DCP and 47% and 75%, respectively, for AFP. A combination of these markers (AFP, AFP-L3, DCP) only marginally improves the accuracy for early detection of HCC [Carr et al. 2007]; the combination of tests may increase sensitivity but the specificity for detecting early HCC declines.

How often should HCC surveillance be performed?

The recommended interval between HCC surveillance tests is 6–12 months. This interval is based on the median doubling time of HCC, which is estimated to range between 80 and 117 days [Kubota et al. 2003; Okada et al. 1993; Sheu et al. 1985; Yoshino, 1983]. One study reported no significant differences in receipt of treatment or survival between HCC patients who receive semiannual compared with annual HCC surveillance [Trevisani et al. 2002].

Recall strategy following an abnormal HCC-surveillance test

Once a surveillance test is abnormal, the most reliable diagnostic tests are triple-phase helical CT and triple-phase, dynamic contrast-enhanced MRI; the latter is slightly better in the characterization and diagnosis of HCC. The hallmark of HCC during CT or MRI is the presence of arterial enhancement, followed by delayed hypointensity of the tumor in the portal venous and delayed phases. Diagnosis of HCC in a patient with cirrhosis (i.e. high pretest probability of HCC) can be confidently established if a focal hepatic mass >2 cm is identified with CT or MRI technique that shows typical features (Figure 1). Generally, both tests are required to be indicative of HCC to avoid biopsy. For focal hepatic mass with atypical or discrepant (between CT and MRI) imaging findings, a mass size of 1–2 cm, or a focal hepatic mass detected in a noncirrhotic liver should undergo a biopsy. A negative biopsy result does not completely rule out malignant disease; and the nodule should be further studied at 3- to 6-month intervals until it is seen to disappear, enlarge, or display diagnostic characteristics of HCC. Nodules smaller than 1 cm should be followed with ultrasound repeated at 3- to 6-month intervals. If, over a period of 2 years, growth has not been observed, routine surveillance at 6-month intervals is suggested.

Figure 1.

Recall strategy following an abnormal hepatocellular carcinoma (HCC)-surveillance test in a patient with cirrhosis (adapted from the United States Department of Veterans Administration guidelines). AFP, Alpha fetoprotein.

Efficacy of HCC surveillance: a review of the evidence

There is intermediate-strength evidence for the efficacy of HCC surveillance in HBV-infected patients. Mixed findings were reported from the two randomized controlled trials conducted in China among individuals with chronic HBV infection with and without cirrhosis. In one placebo-controlled randomized study of nearly 19,000 HBV-infected patients, it was shown that HCC surveillance with both abdominal ultrasound and serum AFP repeated at 6-month intervals resulted in a 37% reduction in HCC-related mortality [Zhang et al. 2004]. In the other randomized controlled trial of 5581 HBV-infected patients, it was reported that serum AFP repeated at 6-month intervals did not result in a significant reduction in the overall mortality [Chen et al. 2003]. The latter study was smaller and has detected HCC less frequently than the first study and therefore could have suffered from a lack of adequate statistical power. Given the emergence of guidelines recommending surveillance, a future randomized controlled trial of HCC surveillance versus placebo is unlikely to be conducted. In addition, these two randomized trials, a population-based surveillance program using serum AFP every 6-months among HBV-infected patients in Alaska reported that during a 16-year follow-up period, patients with HCC who received surveillance had significantly longer 5-year survival among compared with historical controls (42% vs. 0%, respectively) [McMahon et al. 2000].

There are no randomized controlled trials for HCC surveillance in HCV-infected patients or in patients with cirrhosis. Several nonrandomized trials and observational cohort and case–control studies have reported that patients who undergo HCC surveillance are diagnosed at an earlier stage of HCC, are more likely to receive potentially curative therapy, and have a significant reduction in cancer-specific mortality compared with patients detected with symptomatic HCC [Trevisani et al. 2004, 2002; Yu et al. 2004; Wong et al. 2000; Yuen et al. 2000]. For example, in one retrospective cohort study of 680 patients from Taiwan with HBV or HCV infection irrespective of cirrhosis who developed HCC, receipt of routine or opportunistic (for incidental or nonhepatic purposes) ultrasound was associated with a 63% reduction in mortality compared symptomatic diagnosis only [Yu et al. 2004]. However, the positive results reported by these observational studies must be interpreted in the context of almost unavoidable potential biases such as lead time and length bias.

Cost-effectiveness of HCC surveillance

Several studies have used a Markov decision model to evaluate the cost-effectiveness of HCC surveillance. One cost-effectiveness analysis among patients with HCV-related cirrhosis found that annual ultrasound combined with biannual AFP provides the most quality-adjusted life-year gain while still maintaining a cost-effectiveness ratio less than US$50,000/quality-adjusted life-year [Lin et al. 2004]. AFP and ultrasound every 6 months has also been shown to increase quality-adjusted life-years in patients who developed HCC who received resection or liver transplant [Patel et al. 2005]. Although CT scan has not been tested for HCC surveillance purposes, it has also been reported as cost-effective in transplant-eligible patients [Arguedas et al. 2003]. Cost-effectiveness of HCC surveillance is largely dependent on the receipt of potentially curative therapy. If treatment is not available, not given, or the patient is ineligible for treatment due to liver disease severity or other medical comorbidities, then HCC surveillance cannot be considered cost-effective.

Utilization of HCC surveillance in clinical practice

The extent of utilizing HCC surveillance in clinical practice is low. In a recent study of 13,002 HCV-infected veterans diagnosed with cirrhosis during 1998–2005, only 12% received annual surveillance in the 3 years following their cirrhosis diagnosis, and less than 50% received a surveillance test in the first year following cirrhosis diagnosis [Davila et al. 2010a]. At least three retrospective studies among patients with newly diagnosed HCC also found very low rates of surveillance prior to their HCC diagnosis [Davila et al. 2010b, 2007; Leykum et al. 2007]. In the largest of these studies, we reported that among 541 patients diagnosed with HCC during 1994–2002 who had a prior diagnosis of cirrhosis, only 29% received annual surveillance in the 3 years before HCC diagnosis [Davila et al. 2010b].

The low utilization of HCC surveillance likely reflects a combination of cognitive (knowledge of guidelines, new therapeutic options) as well as logistical factors, such as the need for repeated testing over relatively short periods of time, somewhat complicated diagnostic evaluation for HCC, and the limited availability of liver transplant centers to refer patients who are diagnosed with HCC. These factors are likely obstacles facing the implementation of any effective HCC surveillance program.

Recommendation

HCC surveillance should be considered in patients with cirrhosis or advanced hepatic fibrosis irrespective of etiology, and in adult patients with HBV irrespective of cirrhosis. Careful assessments should be undertaken for overall medical and psychological status because these factors determine future potential eligibility for curative or palliative therapies. Patient preferences should be discussed and documented. Based on the evidence presented above, a combination of liver ultrasound and serum AFP for HCC surveillance is recommended every 6 months although a 1-year interval may be equally effective. Consideration should be given to establishing local protocols for surveillance, recall, and diagnosis followed by referral to centers of excellence in the management of HCC.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest statement

None declared.