Risk Factors for Hepatocellular Carcinoma in a Cohort Infected With Hepatitis B or C

Scott R Walter; Hla-Hla Thein; Heather F Gidding; Janaki Amin; Matthew G Law; Jacob George; Gregory J Dore


J Gastroenterol Hepatol. 2011;26(12):1757-1764. 

In This Article


This study identifies sociodemographic and health factors predictive of developing HCC among a cohort of people with chronic HBV or HCV infection in NSW. The incidence of HCC increased with age and comorbidity score, and was higher among males, metropolitan residents, and those with ALD, and particularly cirrhosis. Older age, being male, and having a high comorbidity score were significantly and independently associated with the risk of HCC. Co-infection with HBV and HCV was associated with increased HCC risk in the HCV cohort, and of all risk factors considered, cirrhosis conferred the greatest additional risk, regardless of infection type.

The risk of HCC was observed to increase significantly with age for both HBV- and HCV-infected groups. Such an increase in risk with age has been widely observed,[32] but in some countries, age-specific incidence peaks in the 60s, rather than late 70 s or beyond, possibly due to variation in the prevalence of certain risk factors between regions.[17,21]

Being male is another well-known risk factor for HCC, although there is considerable regional variation in the relative risk compared to females.[21] While relative risks of males compared to females for HCC among the general population range from close to one to almost nine, in most regions in the world, males have two to four times the risk of females,[21] consistent with our results.

A clear increase in the risk of HCC with comorbidity score was observed for both infection groups. Few studies have assessed the association between comorbidity score and HCC risk, as it is often more informative to examine individual health conditions. Each comorbid condition influences the risk of HCC by varying degrees, and multiple conditions might interact in complex ways. However, in terms of identifying high-risk individuals, the comorbidity score quantifies the combined effect of multiple conditions, without the need to interpret risks associated with multiple factors and their interactions.

A number of papers have reported a twofold to threefold increase in HCC risk due to diabetes,[14–16,26] some of which also found an interaction between viral hepatitis and diabetes.[14,16] While including diabetes in the models suggested an increase in the risk of comparable magnitude, the difference was not sufficiently significant to remain in the final model, suggesting that in our cohort, this is a low-risk condition relative to other factors considered.

Alcohol consumption has been identified as a key risk factor for HCC, interacting synergistically with chronic viral hepatitis infection,[16,33] but relatively few studies have examined the risk associated with ALD.[34] We observed significant risk associated with hospitalization with this condition, possibly through the combined effects of alcohol-related and hepatitis-related liver injury. The observed increased risk of HCC among those with HBV/HCV co-infection is consistent with other studies, which found that the combined effect of the two infections is more than additive, but less than multiplicative.[17,18,20,35]

Cirrhosis is well known as the precursor for the vast majority of chronic viral hepatitis-related HCC cases.[19,36,37] Not surprisingly, our study identified cirrhosis as the strongest predictor of HCC for both HBV and HCV cohorts. Two studies based in Taiwan found a 12-fold and 50-fold increase in risk due to cirrhosis among a HBV-infected cohort.[25,38] Sherman reports a more than 20-fold increase in HCC incidence in people with HCV and cirrhosis, compared to those with HCV alone.[36] The magnitude of these estimates approximately agrees with the very high risk identified in our study. The combined effects of cirrhosis and ALD indicated further amplified risk of HCC, particularly among those with HCV. Having a hospital record for both conditions likely indicates advanced or rapidly-progressing liver disease.

A limitation of this study was the incompleteness of country of birth information in the viral hepatitis notification data, which inhibited analysis of the differential risk of developing HCC between people born in different regions. This is particularly pertinent, given that more than half of the HBV-infected group have immigrated from HBV-endemic countries, such as China and Vietnam,[39,40] while the majority of those with HCV are Australian born.[41] Region of birth might also confound the association between remoteness and HCC, since there are much higher proportions of Asian born people in metropolitan areas than non-metropolitan areas.[42] This is particularly likely to be a factor among the HBV-infected cohort. More limited access to specific HCC diagnostic services in non-metropolitan areas might also be a factor in producing an apparently lower incidence of HCC.

A further limitation was the availability of cirrhosis data only through hospitalization codes, particularly as liver biopsy and hepatic elastography diagnosis are generally undertaken through outpatient services. Also, linked treatment data were not available for this study, which eliminated the possibility of examining the extent to which antiviral therapy reduces HCC risk; however, this might form the basis of future studies when additional data permit.

HCC screening and surveillance among at-risk groups have only relatively recently been shown to improve survival.[43,44] Cases only presenting when symptomatic often have a poorer prognosis and fewer treatment options than those detected in the asymptomatic stage of disease.[24,45] In light of its ability to detect tumors early and improve treatment eligibility and survival, screening and surveillance play a key role in reducing the burden of HCC. However, it is only practiced by some groups in NSW, with less than 20% of HCC cases being identified via surveillance.[45,46] Identifying and quantifying risk factors specific to a population, as we have done, forms an integral part of targeting cost-effective surveillance and provides motivation for more widespread screening of high-risk groups.

Antiviral therapy has been shown to limit the progression of liver disease, and in some HBV cases, can reverse decompensated cirrhosis, considerably reducing the risk of HCC.[37,47] Thus, antiviral therapy of chronic viral hepatitis represents a pivotal pathway for reducing the burden of HCC. This also bolsters the case for increasing treatment uptake in general among those with chronic viral hepatitis infection, given that currently, approximately 5% of HBV-infected people[48] and 1–2% of HCV-infected people[41] receive antiviral therapy. A combination of surveillance and treatment has been shown to be a more cost-effective way to reduce the burden of liver cancer than surveillance alone.[46]

In summary, this study has identified and quantified important risk factors for HCC within a high-risk, population-based cohort. Several key factors emerged as independent and significant risks for HCC. Although some previously-reported risk factors were not significant in our analysis, those that were identified were largely consistent with studies conducted in other regions of the world. The association with older age highlights the potential impact of HBV and HCV screening of at-risk groups and early clinical assessment. Antiviral therapy for chronic viral hepatitis is an important strategy for preventing HCC, and further research is required to quantify its mitigation of HCC risk at a population level in the Australian context.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.