Hepatitis B virus maintains its pro-oncogenic properties in the case of occult HBV infection
January 2004, Volume 126, Number 1
Teresa Pollicino, Unit of Clinical and Molecular Hepatology, Department of Internal Medicine, University of Messina, Messina, Italy
This past week I emailed to you (all reports are posted & archived on the NATAP website) finding that occult HBV was not associated with faster liver disease progression in their study of HCV+ IDUs. Here is a study looking at occult HBV & an association with the development of liver cancer (HCC).
"…HBV has been classified as a group 1 human carcinogen and is considered the second most important oncogenic agent after smoking tobacco…Our results indicate that cryptic HBV infection is a risk factor for development of HCC… when occult HBV infection coexists with other causes of liver disease (HCV, alcohol abuse, iron overload, and so on), it might be a cofactor in the pathogenesis of the hepatic injury…testing HBsAg-negative individuals with progressive chronic hepatitis or frank cirrhosis for HBV DNA seems to be of primary importance for the identification of patients with a higher risk of developing liver cancer…"
Background & Aims: Occult hepatitis B virus (HBV) infection is characterized by persistence of HBV DNA into the tissue of hepatitis B surface antigen-negative individuals. The clinical relevance of this peculiar infection is still under debate. In particular, the impact of occult HBV infection in cases of hepatocellular carcinoma (HCC) is uncertain. We investigated the prevalence and molecular status of occult HBV in patients with HCC.
Methods: We tested tumor tissues from 107 patients with HCC and the corresponding nontumor liver tissue from 72 of these patients for HBV DNA. We also examined liver specimens from 192 patients with chronic hepatitis. All cases were hepatitis B surface antigen negative. Covalently closed circular HBV genomes, HBV transcripts, and viral integrated forms were investigated in cases of HCC found positive for occult HBV.
Results: Viral DNA was detected in 68 of 107 cases of HCC (63.5%) and in 63 of 192
cases of chronic hepatitis (32.8%) (P < 0.0001; odds ratio, 3.6; 95% confidence interval, 2.2--5.9). The significant association of occult HBV with HCC was irrespective of age, sex, and contemporary hepatitis C virus infection. Both integrated viral DNA and covalently closed circular HBV genomes were detected in patients with occult HBV. Moreover, the presence of free HBV genomes was associated with persistence of viral transcription and replication.
Conclusions: Our findings provide clear evidence that occult HBV is a risk factor for development of HCC and show that the potential mechanisms whereby overt HBV might induce tumor formation are mostly maintained in cases of occult infection.
Hepatitis B virus (HBV) infection is a major health problem worldwide; it is one of the 10 leading causes of mortality, with an estimated 400 million people chronically infected at the turn of this millennium. HBV infection may associate with a large spectrum of clinical forms, ranging from very mild and asymptomatic clinical pictures to the most severe liver diseases including fulminant hepatitis, cirrhosis, and hepatocellular carcinoma (HCC). In particular, chronic HBV infection is commonly considered a primary risk factor for development of HCC, exerting its pro-oncogenic properties through both indirect and direct mechanisms. The indirect mechanisms are related to its propensity to induce continuous or recurrent phases of liver necroinflammation and to promote the progression of chronic hepatitis to cirrhosis, which is the step preceding the development of HCC in most cases. Schematically, the direct carcinogenic mechanisms have been
related to the capacity of HBV to integrate into the host's genome and to produce proteins (X protein and truncated preS-S protein) provided of potential transforming properties.
HBV infection is usually diagnosed when the circulating hepatitis B surface antigen (HBsAg) is detected. However, the availability of highly sensitive molecular biology techniques has allowed the identification of HBV infection in HBsAg-negative individuals with or without circulating antibodies to HBsAg and/or hepatitis B core antigen (anti-HBc). Much evidence suggests that this so-called occult or cryptic HBV infection is highly prevalent and may have a relevant clinical impact. In particular, we and others previously showed that occult HBV infection
significantly correlates with cirrhosis in chronic hepatitis C virus (HCV) carriers.
Since the early 1980s, a certain number of studies have reported the detection of HBV DNA in liver tissue of HBsAg-negative patients with HCC. With the advent of the most sensitive polymerase chain reaction (PCR) techniques, this observation has been confirmed; however, extensive studies performed in homogeneous populations and evaluating the real prevalence of this event are still lacking. Another aspect that needs to be highlighted is the molecular status of the occult HBV in patients with HCC, which is an essential support for understanding the possible pathogenic mechanisms through which the cryptic HBV might contribute to hepatocellular transformation.
In this study, we investigated the presence and molecular features of HBV DNA in liver tissues from a large series of HBsAg-negative patients with HCC.
DISCUSSION by authors
HCC is one of the most common cancers worldwide, and its incidence appears to be increasing in the more industrialized areas of North America, Europe, and Asia. Human HCCs mostly occur in patients with long-lasting chronic hepatitis due to infection with HBV and/or HCV. However, while this malignancy is still mainly related to an overt (namely, HBsAg positive) HBV infection in developing countries, the increasing rate of HCC observed in the past decade in the western world seems to be more frequently linked with cirrhosis C. Much evidence
indicates that cryptic HBV infection is a phenomenon often occurring in chronic HCV carriers, and we previously reported that about one third of individuals with type C CLD (chronic liver disease) in our area have a masked HBV coinfection. In the present study, our examination of a large series of liver tissues from HBsAg-negative individuals with HCC showed that most of these patients carry occult
HBV infection, with a significantly higher prevalence compared with that observed in subjects with HBsAg-negative CLD; this significance was also maintained when the comparison was performed with the subgroup of non-HCC cirrhotic patients. Moreover, the statistical analysis showed that occult HBV strongly associated with liver cancer independently of age, sex, and contemporary HCV infection.
Our results indicate that cryptic HBV infection is a risk factor for development of HCC. This is in agreement with data obtained in animal models showing that both woodchucks and ground squirrels, once infected by the corresponding hepadnaviruses (woodchuck hepatitis virus and ground squirrel hepatitis virus, respectively), are at high risk of developing HCC after the apparent
clearance of the virus. The mechanisms responsible for hepatocellular transformation in these rodents are essentially related to the propensity of the viruses to act as an insertional mutagen of the host's myc family genes.
However, it was shown that besides integrating into the host DNA, the viral genomes may also persist as free episomal forms during occult woodchuck hepatitis virus infection; an interesting study on woodchucks convalescent from acute viral hepatitis showed the lifelong persistence of small amounts of replicating virus that induce a mild liver necroinflammation continuing for life.
Also, occult HBV may persist as episomal DNA after HBsAg seroclearance, and it is able to induce mild inflammation persisting for decades. Moreover, when occult HBV infection coexists with other causes of liver disease (HCV, alcohol abuse, iron overload, and so on), it might be a cofactor in the pathogenesis of the hepatic injury. In this context, we have to mention that occult HBV infection seems to favor the progression of chronic hepatitis toward cirrhosis, which is known to be the main precancerous lesion of the liver. However, our present data indicate that occult HBV is associated with HCC independently of cirrhosis. Of note, in our series of patients with HCC developed on cirrhotic livers, we found free viral genomes in
a considerable number of cases. These viruses maintain low levels of transcription and replication, as indicated by the presence of HBV cccDNA and viral messenger RNAs (mRNAs) that, interestingly, were detected in both tumor and nontumor samples. Although we cannot completely rule out possible contamination between tumor and surrounding nontumor liver tissues, several considerations suggest that free replicative forms of HBV were really present in neoplastic cells: (1) all 4 HBV genomic regions were amplified in tumor extracts but not in the corresponding nontumor specimen from several cases; (2) analogously, some cases had HBV cccDNA in tumor but not in the paired nontumor samples; (3) a discrete nucleotide divergence was found between HBV sequences isolated from tumor and nontumor specimens of individual patients; and (4) in one case, viral strains belonging to 2 different genotypes were detected in the tumoral liver whereas only one HBV genotype was shown in the surrounding nontumoral tissue. The sequence divergences observed between viral isolates from tumor and nontumor tissues might reflect the heterogeneity of the HBV infecting the liver in toto and tempts us to speculate on a possible segregation of some viral strains in transformed and clonally expanded cells. In any case, what clearly emerged from our study is that ongoing covert HBV replication frequently occurs in HBsAg-negative cirrhotic patients who develop HCC. The long-term persistence of cryptic yet productive HBV infection might exert an indirect oncogenic role by chronically sustaining a mild necroinflammation that may contribute to the development of cirrhosis and liver cancer. Moreover, occult HBV might play a direct oncogenic role through both its integration into the host genome and the maintained transcriptional activity that, although at low levels, may allow the synthesis of proteins (such as X protein) with potential pro-oncogenic properties. In summary, our data show that the direct and indirect mechanisms whereby overt HBV might induce tumor formation are mostly maintained by occult HBV.
HBV has been classified as a group 1 human carcinogen and is considered the second most important oncogenic agent after smoking tobacco. Our study definitively shows that HBV also maintains its oncogenic role in the case of occult infection. Consequently, testing HBsAg-negative individuals with progressive chronic hepatitis or frank cirrhosis for HBV DNA seems to be of primary importance for the identification of patients with a higher risk of developing liver cancer.
We studied liver specimens from 107 HBsAg-negative cirrhotic patients with HCC (81 men and 26 women; mean age, 63.7 ± 10 years) randomly collected from January 1999 to December 2000 at 7 Italian liver units located in distinct geographic areas of the country. Seventy-three of the 107 patients were antibody to HCV (anti-HCV) positive; among the 34 anti-HCV-negative individuals, 5 had alcoholic liver disease and 29 had cryptogenic liver disease. Serum markers of previous HBV infection were available for 82 of the 107 patients.
Twenty-five of these 82 patients were anti-HBc positive and 11 of them were also antibody to HBsAg positive; the remaining 57 cases were negative for all HBV serum markers. Tumor liver specimens from the 107 patients and the corresponding nontumor tissue from 72 of the patients were available for molecular analysis. Tissue samples had been obtained by surgical resection
or percutaneous needle liver biopsy and immediately frozen and stored at --80°C.
A total of 192 frozen liver biopsy specimens from HBsAg-negative patients (125 men and 67 women; mean age, 48.2 ± 11.5 years) with chronic liver disease (CLD) (143 with chronic hepatitis and 43 with cirrhosis) were available for inclusion in this study as a control group. A total of 153 of the 192 patients had an HCV-related CLD; among the 39 anti-HCV-negative individuals, 7 had alcohol-related liver disease and 32 had cryptogenic liver disease. These specimens were collected from June 1999 to May 2000 at 3 of the previously mentioned Italian centers (Messina, Palermo, and Turin).
Serum markers of previous HBV infection were available for 119 of the 192 patients. Thirty-three of these 119 patients were anti-HBc positive and 18 of them were also antibody to HbsAg positive; the remaining 86 cases were negative for all HBV serum markers. None of these patients had been treated with antiviral or immunosuppressive drugs before undergoing liver biopsy, and none were infected with the human immunodeficiency virus.
Identification of occult HBV infection in liver tissue
We investigated the presence of HBV DNA sequences in liver tissue specimens from 107 HBsAg-negative patients with HCC and 192 HBsAg-negative patients with CLD (chronic liver disease) by nested PCR and the use of 4 primer sets, each specific for S, C, Pol, and X viral genomic regions, respectively. As we previously reported and as also recommended by others, we considered cases in which sequences of at least 2 different HBV genomic regions were detected as occult HBV-positive cases. In this way, 68 of the 107 patients with HCC (63.5%) and 63 of the 192 patients with CLD (32.8%; P < 0.0001) were positive for occult HBV, with a univariate odds ratio for the association between occult HBV and HCC of 3.6 (95% confidence interval, 2.2--5.9).
When we divided the study populations on the basis of anti-HCV positivity, we found HBV DNA in 45 of the 73 anti-HCV-positive patients with HCC (61.6%) and in 56 of the 153 anti-HCV-positive patients with CLD (36.6%; P < 0.001), resulting in an estimated univariate odds ratio of 2.9 (95% confidence interval, 1.6--5.1). Among the anti-HCV-negative patients (34 with HCC and 39 with CLD), we detected viral sequences in the liver of 23 patients with HCC (67.6%) and 7 patients with CLD (18%; P < 0.0001) and the univariate odds ratio was 9.6 (95% confidence interval, 3.2--28.4).
Notably, in the group of patients with HCC, we found no correlation between occult HBV infection and anti-HCV positivity (P = 0.5); in the group of patients with CLD, the presence of intrahepatic HBV sequences was significantly associated with HCV infection (P = 0.03), in accordance with previous data reported by us and others (reviewed by Torbenson and Thomas18). The prevalence of occult HBV infection did not differ with sex or age either among patients with HCC or those with CLD. As expected, the patients with HCC were older (mean age, 63.7 ± 10 years) than those with CLD (mean age, 48.2 ± 11.5 years). Consequently, to verify whether the relationship of occult HBV infection with HCC was affected by age, we performed multivariate analysis by using the Kendall's rank correlation test on the total of 299 cases (107 patients with HCC and 192 patients with CLD). There was a correlation between age and occult HBV ( = 0.2; P = 0.00002), between age and HCC ( = 0.46; P < 0.000001), and between occult HBV and HCC ( = 0.3; P < 0.0001). Of note, occult HBV and HCC were associated regardless of age ( = 0.23; P < 0.000001).
As previously mentioned, all patients with HCC had cirrhosis, as did 43 of the 192 individuals in the control group. HBV DNA was detected in 19 of these 43 subjects (44.2%). The comparison of the prevalence of occult HBV between HCC and non-HCC cirrhotic patients showed the persistence of the significant association between HBV and liver cancer (P = 0.03). Moreover, the Kendall's test showed that occult HBV is associated with HCC independently of cirrhosis ( = 0.18; P = 0.015).
Data on serum anti-HBc reactivity were available for 82 patients with HCC and 119 patients with CLD. Among patients with HCC, occult HBV was detected in 21 of 25 anti-HBc-positive (84%) and 32 of 57 anti-HBc-negative cases (56.1%; P = 0.01). Therefore, not only most of the anti-HBc-positive individuals but also more than half of subjects negative for all serum markers of HBV infection carried occult HBV. The analysis of patients with CLD showed the presence of HBV sequences in 15 of 33 anti-HBc-positive (45.5%) and 22 of 86 anti-HBc-negative patients (25.6%;
P = 0.04). These results are in accordance with data that emerged from a recent review of the literature that, besides the significant association between occult HBV and anti-HBc positivity, also showed that more than 20% of patients with chronic hepatitis and negative for all HBV serum markers prove to be carriers of cryptic HBV infection.