Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related death worldwide and has an increasing incidence in the United States due to the current epidemic of nonalcoholic fatty liver disease and hepatitis C virus. Prognosis for patients with HCC depends on tumor stage at diagnosis, with curative options only available for patients diagnosed at an early stage. Patients with early stage HCC achieve 5-year survival rates near 70% with resection and transplantation, whereas those with advanced stage HCC have a median survival of less than 1 year.[2, 3]
Surveillance at regular intervals is recommended in all patients with cirrhosis to detect HCC at an early stage. Although the most recent American Association for the Study of Liver Diseases (AASLD) guidelines recommend the use of ultrasound alone for HCC surveillance, there has been extensive debate regarding the use of alpha-fetoprotein (AFP) in clinical practice.[4-6] The removal of AFP from the AASLD guidelines was questioned in a study demonstrating that AFP may be beneficial for surveillance in clinical practice given the gap between the efficacy and effectiveness of ultrasound.[6, 7] Although ultrasound has a pooled sensitivity of 63% for detecting HCC at an early stage in prospective cohort studies (Fig. 1), its sensitivity in clinical practice is substantially lower at 32%. The sensitivity of using ultrasound and AFP in combination was 63.4%, significantly higher than that of ultrasound alone (P < 0.001), with minimal loss in specificity. The variable effectiveness of ultrasound may be related to differences in operator experience and technique, with many patients in the United States receiving their ultrasound examinations in local community centers instead of tertiary care centers. Furthermore, the ability of ultrasound to accurately visualize the liver in patients with morbid obesity or a very nodular liver may be impaired. In contrast to current guidelines, these studies suggest that AFP may play an important role in HCC surveillance among patients with cirrhosis.
It is clear that better surveillance tools, including more accurate biomarkers or more cost-effective advanced imaging with low or no risk of radiation, are necessary to help improve the sensitivity of finding tumors at an early stage. Des-γ carboxyprothrombin (DCP) and lectin-bound AFP (AFP-L3%) are two potential biomarkers that have been proposed. In a nested case-control study among patients in the Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis (HALT-C) Trial, DCP and AFP had sensitivities of 74% and 61%, respectively, for HCC at any stage, which was increased to 91% by using the two markers in combination. However, AFP was demonstrated to be more sensitive than DCP and AFP-L3% for the diagnosis of early HCC in a large, multicenter case-control study. Further studies are necessary to better evaluate the role of new biomarkers, including AFP-L3% and DCP, prior to their routine use in clinical practice.
Some studies have proposed that computed tomography (CT) or magnetic resonance imaging (MRI) may be more sensitive as alternative imaging studies for the detection of HCC. A systematic review found that the sensitivity of CT for HCC at any stage was 68% (95% confidence interval [CI], 55%-80%) and that of MRI was 81% (95% CI 70%-91%). Although these numbers are encouraging, CT and MRI have only been studied as diagnostic tests and not as surveillance tests with regard to early stage HCC in patients with cirrhosis. Additionally, the increased cost and potential adverse effects (such as radiation exposure) limit their use in surveillance. There is currently insufficient evidence for the use of CT or MRI in routine clinical practice. Ultrasound, with or without AFP, remains the recommended surveillance strategy until further studies are performed.
The best evidence supporting HCC surveillance comes from a large randomized controlled trial in China among more than 19,000 patients with chronic hepatitis B infection, although this study has been criticized for violating intention-to-treat principles and failing to account for cluster randomization in the analysis. Patients randomized to surveillance had significantly better tumor stage (60.5% versus 0% stage I tumors), higher rates of curative treatment (46.5% versus 7.5%), and a 37% reduction in annual mortality (P < 0.01) (Fig. 2). Although there has never been a randomized controlled trial among patients with cirrhosis, there have been several prospective cohort studies demonstrating that patients undergoing surveillance have earlier stages of disease and better overall survival than patients with cirrhosis who had not undergone surveillance. When a randomized trial among patients with cirrhosis was attempted, it was deemed to be not feasible given that both the patients and their physicians believed prior studies provided sufficient evidence for HCC surveillance.
HCC surveillance is currently recommended for all patients with cirrhosis, regardless of other risk factors, although the risk of developing HCC is likely not uniform across all patients. A large registry-based study from Denmark suggested that some patients, such as those with alcoholic cirrhosis, have a low risk for HCC and may not truly benefit from HCC surveillance. Accurate assessment of HCC risk among patients with cirrhosis may allow targeted application of HCC surveillance programs, given that HCC surveillance is only cost-effective among those with an annual HCC risk of >1.5%. Several retrospective case-control studies have previously identified risk factors for HCC among patients with cirrhosis, including older age, male sex, diabetes, and alcohol intake. A recent predictive model was able to identify patients at high risk for HCC development with moderate accuracy (c-statistic, 0.76; 95% CI, 0.67-0.85). However, further refinement of predictive models in larger data sets is still necessary prior to routine use in clinical practice.
Although surveillance is efficacious for detecting early HCC, its effectiveness in clinical practice is determined by several factors, including low rates of use.[17-19] A recent meta-analysis demonstrated that <20% of patients with cirrhosis undergo surveillance. Although screening rates are higher among patients receiving subspecialty care, only 20% to 40% of patients with cirrhosis are followed by gastroenterologists or hepatologists nationally. Underuse of HCC surveillance is related to multiple failure points, including 40% of patients having unrecognized liver disease and/or cirrhosis prior to HCC presentation; however, the most common reason for lack of HCC surveillance is physicians failing to order HCC surveillance in patients with known cirrhosis (Fig. 3). Downstream issues, such as radiologic capacity and patient acceptance, do not appear to be major barriers to HCC surveillance at this time. A secondary analysis of the HALT-C Trial suggested that these failures in the surveillance process potentially contribute to more advanced tumor stage at presentation in up to one-third of patients.
In conclusion, HCC surveillance is recommended every 6 months in patients with cirrhosis to detect tumors at an early stage. However, <20% of patients with cirrhosis undergo HCC surveillance, contributing to high rates of advanced tumor stage at presentation. Although better surveillance tools are needed to improve the sensitivity of finding tumors at an early stage, there is currently insufficient evidence for the routine use of cross-sectional imaging or other biomarkers, such as AFP-L3 and DCP. Given the variable effectiveness of ultrasound in clinical practice, surveillance should be performed using a combination of ultrasound and AFP at this time to optimize the detection of tumors at an early stage.