Predictive Scores for Hepatocellular Carcinoma Development in Chronic Hepatitis B Virus Infection: “Does One Size Fit All?”
Department of Hepatology, Institute of Liver and Biliary Sciences, New Delhi, India published online 22 February 2012.
Yang HI, Yuen MF, Chan HL, et al. Risk estimation for hepatocellular carcinoma in chronic hepatitis B (REACH-B): development and validation of a predictive score. Lancet Oncol 2011;12:568–574.
The study by Hwai-I Yang et al from Taiwan is an important study that sought to develop and validate a scoring system to predict the risk of hepatocellular carcinoma (HCC) development in patients with chronic hepatitis B virus (HBV) infection (Lancet Oncol 2011;12:568–574). The development cohort consisted of patients from the population-based prospective REVEAL-HBV database of 3584 hepatitis B surface antigen–positive patients aged 30–65 years with a median follow-up of 12.0 years (131 patients developed HCC). The validation cohort consisted of 1505 patients from 3 hospitals in Hong Kong and South Korea, with a mean follow-up of 7.3 years (111 patients developed HCC). None of the patients received antiviral therapy in either of the groups. Variables included in the risk score were gender, age, serum alanine aminotransferase (ALT) concentration, hepatitis B e antigen (HBeAg) status, and serum HBV DNA level. Cox multivariate proportional hazards model was used to predict risk of HCC at 3, 5, and 10 years.
Most of the patients were HBeAg negative (84.4% and 61.6% in development and validation cohorts, respectively) and noncirrhotic at baseline (0% and 82% in development and validation cohorts, respectively). An empirical 17-point risk score was developed with HCC risk ranging from 0.0% to 23.6% at 3 years, 0% to 47.4% at 5 years, and 0.0% to 81.6% at 10 years for patients with the lowest and highest HCC risks, respectively. The area under receiver operating characteristic curves to predict risk were 0.811 (95% confidence interval [CI], 0.790–0.831) at 3 years, 0.796 (95% CI, 0.775–0.816) at 5 years, and 0.769 (95% CI, 0.747–0.790) at 10 years in the validation cohort, and 0.902 (95% CI, 0.884–0.918), 0.783 (95% CI, 0.759–0.806), and 0.806 (95% CI, 0.783–0.828), respectively, after the exclusion of 277 patients with cirrhosis in the validation cohort.
Back to Article Outline Comment HCC is a potentially serious complication of chronic HBV infection. Approximately 75% of HCC cases occur in patients with cirrhosis, although HCC can develop in noncirrhotics as well. Overall cohort studies show that the incidence of HCC was <0.01%–0.2% per year in so-called inactive hepatitis B surface antigen carriers, <0.3%–0.9% per year in chronic HBV, and 1.5%–6.0 % in cirrhosis (J Hepatol 2008;48:335–352). Chronic HBV infection acquired in adults has a much lower risk of HCC development than infections resulting from early life transmission, possibly reflecting differences in duration of infection.
Several potential factors have been identified to be associated with a higher risk of development of HCC. Complex interactions between viral factors, patient factors including genetic factors, and other environmental factors led to HCC development in chronically HBV-infected patients.
A significant dose–response association between serum HBV DNA levels at initial evaluation and subsequent risk of HCC has been demonstrated. Recently, it has been shown that patients with similar HBV DNA levels at baseline have different risks of HCC depending on their HBV DNA levels during follow-up; greater decreases in serum HBV DNA during the follow-up being associated with lower HCC risk (Gastroenterology 2011 [Epub ahead of print]).
The risk of HCC has been found to be higher for patients infected with genotype C (compared with A or B), genotype D (compared with A, in India), and genotype F (in Alaska natives). Risk of HCC may differ among sub genotypes, for example risk of HCC is more in HBV/C2 compared with HBV/C1 (J Gastroenterol Hepatol 2011;26[Suppl 1]:123–130).
Many cross-sectional and case–control studies have consistently demonstrated that basal core promoter A1762T/G1764A mutation is correlated with HCC development and the risk is observed for both genotypes B and C (Gut 2008;57:91–97). Mutations in Pre-S have been recently reported in HCC cases compared with chronic or asymptomatic cases.
Obesity, type 2 diabetes, aflatoxin B1 exposure, and environmental polyaromatic hydrocarbon exposure may also enhance the hepatocarcinogenicity of chronic HBV infection. Familial predisposition is also a risk factor for HCC in chronic HBV infected subjects. Male gender, advancing age, and an elevated ALT level, and persistently elevated alpha-fetoprotein (AFP) levels are also risk factors for the development of HCC. Advanced fibrosis and severe lobular activity on histology at presentation of chronic hepatitis B are independent predisposing risk factors for the development of HCC.
Many other groups have also evaluated the use of prediction models for HCC in patients with chronic HBV. Yuen et al derived a prediction score based on gender, age, HBV DNA levels, core promoter mutations, and the presence of cirrhosis, and found a sensitivity of >84% and specificity of >76% to predict the 5- and 10-year risks for the development of HCC. Cirrhosis was seen in 15.1%, HBeAg positivity in 43.4%, and ALT <1 × the upper limit of normal (ULN) in 50% of patients; 10.7% of patients were treated (J Hepatol 2009;50:80–88).
Another study from Hong Kong, China, developed a predictive score (ranging from 0 to 44.5) based on age, albumin, bilirubin, HBV DNA, and presence of cirrhosis, which was validated in a validation cohort. In the training cohort, 14.2% and 10.4% in the validation cohort had ALT <1 × ULN. Of these patients, 15.1% and 25% received antiviral treatment in the training and validation cohort, respectively (J Clin Oncol 2010;28:1660–1665).
Another recent study from Taiwan, using data from REVEAL (Two-thirds of the Risk Evaluation of Viral Load Elevation and Associated Liver Disease/Cancer–Hepatitis B Virus [REVEAL-HBV] study cohort was allocated for model derivation [n = 2435], and the remaining third was allocated for model validation [n = 1218]), developed a nomogram using gender, age, family history of HCC, alcohol consumption, ALT level, HBeAg status, serum HBV DNA level, and HBV genotype. All area under receiver operating characteristic curves for the risk prediction nomogram were ≥0.82. Most of the participants were noncirrhotics, all were treatment naïve, >80% were HBeAg negative, and >90% had an ALT <1 × ULN (J Clin Oncol 2010;28:2437–2444).
This study by Yang et al included gender, age, serum ALT concentration, HBeAg status, and serum HBV DNA level for calculating the risk score. Although this study used readily available risk parameters for calculating the risk score, many important risk factors have not been included. AFP was not included because testing has variable sensitivity and specificity for the diagnosis of HCC. However, it may be useful in predicting development of HCC, the annual HCC incidence in HBV patients with elevated AFP being increased compared with patients with normal AFP (Hepatology 1994;19:61–66). Genotype and core promoter or precore mutations are also important risk factors; however, they are not widely available as yet. As genotyping and mutation analysis become more widely available, these factors will gain prominence and risk scores incorporating these parameters will be needed.
This risk score was developed and validated in patients from the Far East. This score may not be applicable to patients of other ethnicities and geographical areas, who might be different with regard to age at infection (perinatal vs adult), genetic background, HBV genotype or species, and exposure to environmental factors such as aflatoxin and alcohol. Thus, further validation and the need for newer scores are justified for patients of other ethnic origins.
Most of the patients in this study were HBeAg negative. It has been well documented that patients with HBeAg-negative infection have flares (Gastroenterology 2009;136:1272–1280). There is the likelihood of more hepatocellular injury and chances of development of cirrhosis and HCC among those who have persistently elevated ALT levels compared with those who have intermittently elevated or normal ALT levels. These issues need to be considered in a model. Furthermore, whether this score will be applicable to HBeAg-positive subjects remains to be seen. This study included patients aged ≥30 years, so it may not be applicable to younger patients, who do have the potential for developing HCC (Cancer Sci 2009;100:2249–2254). Co-occurrence with hepatitis C virus, HIV, diabetes mellitus, or alcohol use may affect liver disease progression in patients with chronic hepatitis B, and this predictive score may have limited applicability in patients with these comorbidities.
Almost all patients included in this analysis did not have cirrhosis; therefore, they are considered to be a low-risk population (although a proportion of patients with subclinical cirrhosis may have been missed); therefore, this risk calculator does not apply to cirrhotics. In patients with chronic HBV, advanced liver fibrosis and cirrhosis are significantly correlated with the risk of HCC development. To date, liver biopsy had been the gold standard for assessing the severity of liver fibrosis and cirrhosis. Recently, a liver stiffness measurement using Fibroscan has been introduced. In a recent study that investigated the usefulness of liver stiffness measurement as a predictor of HCC development in patients with chronic hepatitis B, it was found that patients with greater liver stiffness measurement at baseline were at a significantly greater risk of HCC development (Hepatology 2011;53:885–894). Incorporation of this noninvasive test in predictive scores would be highly desirable.
The present score was generated from a natural history cohort using baseline data and without antiviral therapy; whether this risk stratification is suitable to use for changing risk profiles during follow-up either spontaneously or as a result of antiviral treatment is questionable and requires further validation. Treatment with antiviral agents and the resultant lowering of HBV DNA and ALT levels, and HBeAg seroclearance and seroconversion, can change the risk profile of an individual patient. In addition, a recent analysis of REVEAL data has shown that long-term changes in serum levels of HBV DNA and ALT are independent predictors of risk for HCC (Gastroenterology 2010;138:1747–1754). It will be important to establish whether lowering the risk score by successful anti-HBV treatment therapy eventually results in a reduction of future HCC risk. Thus, the ability for risk stratification of patients with HBV in predicting HCC is important from a public health perspective, resource allocation in health care, and individual patient care (including need for enhanced surveillance and need for antiviral treatment). Although the authors are to be complimented for their earnest attempt, there is clearly a need for developing predictive scores for different phases of chronic HBV infection and stages of liver disease. It seems logical that different risk scores for different stages of chronic HBV infection would be more precise and helpful for clinical practice and policy decisions. It needs to be emphasized that the terms like “HBV carrier” and “inactive HBV carrier” are misleading and should best be avoided; chronic, low-level viremia does have a potential for development of HCC (Gastroenterology 2008;134:1376–1384). A better and more widely acceptable term for different phases of HBV infection is chronic HBV infection, with low or high levels of viremia (Gastroenterology 2008;134:1376–1384). Predictive scores also need to be developed in different ethnic groups and geographical areas, which may be different with regard to age at infection, genetic background, HBV genotype or other variants, and exposure to environmental factors. Liver stiffness, an easily measurable parameter, should be incorporated in future predictive score development. Predictive scores often incorporate low-cost, commonly available test parameters for wider applicability; however, with the ready availability of HBV genotypic variant analysis and genome-wide association studies in the future, the combination of genetic and nongenetic factors may promise a more personalized approach to predicting HCC in chronically HBV-infected patients.
|
发表于 2012-4-6 17:50
