Occult hepatitis B virus infection is associated with the development of hepatocellular carcinoma in chronic hepatitis C patients




Occult hepatitis B virus (HBV) infection frequently occurs in patients with HBV surface antigen (HBsAg)-negative chronic liver disease, and much evidence suggests that it is a risk factor for hepatocellular carcinoma (HCC) development. However, to the authors' knowledge, no follow-up study has been performed to date evaluating HCC occurrence over time in chronic hepatitis patients with or without occult HBV infection.


A cohort of the 380 HBsAg-negative chronic hepatitis patients attending the study institution between 1991–2000 were evaluated and tested for occult HBV DNA by analysis of liver biopsy specimens.


There were 135 patients (35.5%) with occult HBV and 245 patients (64.5%) without occult HBV. Cirrhosis was significantly associated with occult HBV infection (P = 0.01). One hundred thirty-four of these patients were followed for a minimum of 50 months (median, 82.8 ± 32.6 mos). Fifty-three patients (39%) were occult HBV carriers and 81 (61%) were not. Nine patients developed HCC during the follow-up; eight were positive and one was negative for occult HBV (P = 0.002).


The current observational cohort study showed that, among the HBsAg-negative patients with chronic hepatitis, HCC develops for the most part in carriers of occult HBV. Therefore, the evaluation of HBV genomes in chronic hepatitis patients appears to be a powerful tool for the identification of individuals at higher risk of HCC development. Cancer 2006. © 2006 American Cancer Society.

Occult hepatitis B virus (HBV) infection is characterized by the persistence of HBV DNA in the liver tissue (± serum) in hepatitis B surface antigen (HBsAg)-negative individuals.1–3 Much recent evidence indicates that this peculiar form of HBV infection is a quite frequent occurrence not only in individuals with circulating antibodies to HBsAg (anti-HBs) and/or hepatitis B core antigen (anti-HBc), but also in those who are negative for all HBV markers.4–9 In particular, occult HBV infection is highly prevalent in chronic carriers of hepatitis C virus (HCV),10–12 and we previously reported that approximately one-third of individuals with type C chronic liver disease in our area have a masked HBV coinfection.12, 13 In these cases occult HBV might have a relevant clinical impact favoring or accelerating the progression of liver fibrosis and the establishment of cirrhosis.10, 12–15 In addition, since the early 1980s occult HBV infection has been suspected of being a possible risk factor for hepatocellular carcinoma (HCC) development.7, 16–19 We recently confirmed this hypothesis by examining the liver tissues from a large series of HBsAg-negative patients with HCC for HBV DNA, and we also showed that occult HBV maintains the potential pro-oncogenic properties typical of the overt (namely, HBsAg-positive) HBV infection.14 However, all the available data in this field come from cross-sectional evaluation of patients with liver cancer, whereas to our knowledge, no follow-up study has been performed to data evaluating the occurrence of HCC development over time in HBsAg-negative chronic hepatitis patients with or without occult HBV infection.

In the current study, we report the results of an observational study evaluating HCC occurrence in a large cohort of subjects who had been tested for occult HBV when undergoing liver biopsy for chronic liver disease.


Fragments of liver specimens from 380 patients with HBsAg-negative chronic liver disease (246 males and 134 females with a mean age of 48.8 ± 12.3 yrs; 154 patients were anti-hepatitis B core [HBc] positive and 226 were anti-HBc negative) (Table 1) who consecutively underwent fine-needle biopsy between January 1991 to December 2000 in the study liver unit were tested for occult HBV DNA through the methods detailed previously.12, 13, 20 Briefly, DNA was extracted from the frozen liver specimens of each case by standard procedures12, 13, 20 and analyzed for the presence of HBV genomes by performing four different in-house single-step or nested polymerase chain reaction (PCR) amplification assays to detect the preS-S, precore-core, Pol, and X viral region, respectively. We considered the cases that demonstrated positivity in at least two different viral genomic regions to be HBV DNA-positive. Appropriate negative controls were included in each PCR. Moreover, direct sequencing of all amplified HBV sequences confirmed the specificity of the reactions. The limit of sensitivity of our single-step and nested PCR methods was in the range of 103 and 10 genome equivalents/mL, respectively. The four sets of primers used for viral DNA amplification (HBV1–HBV4 for the preS–S region, HBV5–HBV8 for the precore–core region, HBV9–HBV12 for the Pol region, and HBV13–HBV16 for the X region) were complementary to conserved regions of HBV genotype D at the following nucleotide position numbered from the EcoR1 unique restriction site of the viral genome (from 5′ to 3′): HBV 1, 2815–2834; HBV 2, 690–671; HBV 3, 2932–2951; HBV 4, 459–440; HBV 5, 2021–2040; HBV 6, 2464–2448; HBV 7, 2048–2066; HBV 8, 2385–2366; HBV 9, 2414–2433; HBV 10, 174–155; HBV 11, 2457–2473; HBV 12, 2951–2932; HBV 13, 1266–1286; HBV 14, 1628–1608; HBV 15, 1380–1400; HBV 16, 1540–1520.

Table 1. Demographic Characteristics, anti-HBc Status, and Histology of the Patients Tested for Occult HBV
Total no. of patients380a
  • HBc: hepatitis B core; HBV: hepatitis B virus; SD: standard deviation.

  • a

    326 hepatitis C virus (HCV)-positive patients and 54 HCV-negative patients (6 with primary biliary cirrhosis, 4 with hemochromatosis, 9 alcoholic-related liver disease cases, 5 drug-related liver disease cases, and 30 cryptogenic liver disease cases).

Mean age in yrs (± SD)48.8 ± 12.3
Male/female ratio246/134
Anti-HBc positive/negative154/226
Histologic diagnosis 
 Minimal changes63
 Chronic hepatitis251

Of the total of 380 patients, 326 were anti-HCV positive, and the HCV genotype was available in 231 of them (genotype 1 in 142 cases, genotype 2 in 59 cases, and genotype 3 in 30 cases). Among the 54 anti-HCV-negative individuals, 9 had alcoholic liver disease, 6 had primary biliary cirrhosis, 4 had hemochromatosis, 5 had drug-related hepatotoxicity, and 30 had cryptogenic liver disease. None of the patients were infected with the human immunodeficiency virus.

The histologic evaluation demonstrated minimal or nonspecific changes in 63 cases, chronic hepatitis in 251 cases, and cirrhosis in 66 cases (Table 1). None of the patients had been treated with antiviral or immunosuppressive drugs before undergoing liver biopsy. Subsequently, 118 of the 326 anti-HCV-positive patients were treated with varying therapeutic schedules including interferon (IFN)-alpha therapy at a dose of 3–6 million units 3 times weekly (93 cases) for 4–12 months, whereas 25 cases were treated with analogous dosages of IFN-alpha plus ribavirin (at a dose of 800–1200 mg/day). Twenty-nine of the former group and eight of the latter group of patients were long-term responders as defined by the conventional criteria.21

After performing liver biopsy, we had the opportunity to follow 134 of the 380 patients (124 of whom were HCV positive and 10 of whom were HCV negative). Each of the individuals followed was examined every 3–6 months at our outpatients' clinic for a minimum of 50 months (median, 82.8 ± 32.6 mos). These 134 patients were significantly older compared with the remaining 246 subjects who were lost to follow-up, whereas the 2 groups did not differ with regard to gender, histologic grade of the liver disease, and HCV genotype distribution (Table 2). The patients followed were unselected, and we have no explanation for justifying the significant difference in the mean age between these patients and those who were not followed. We can only imagine that younger people tend to be less prone to undergo long-term periodic visits in the hospital than older ones.

Table 2. Comparison of Demographic Characteristics, Histology, and HCV Genotype Distribution between Follow-Up Patients and Individuals Lost to Follow-Up
 Follow-up patients (n = 134)Not followed-up patients (n = 246)P
  1. HCV: hepatitis C virus; SD: standard deviation; NS: not significant.

Mean age in yrs (± SD)52 ± 10.648.5 ± 10.120.003
Male/female ratio85/49160/86NS
Histologic diagnosis   
 Minimal changes1746 
 Chronic hepatitis96155NS
Total no. of available HCV genotypes116122 
No. of 1/2/3 genotypes74/30/1271/32/19NS

Statistical Evaluations

Statistical analyses were evaluated using the Student t-test and the chi-square method.


Overall, HBV genomes were found in liver tissue from 135 of the 380 patients (35.5%) tested for occult HBV. In particular, viral DNA was detected in 124 of the 326 (38%) HCV-positive patients and in 11 of the 54 (20%) HCV-negative patients (P < 0.01). Occult HBV was found to be correlated significantly with cirrhosis, and carriers of occult HBV infection were equally distributed in the two groups of patients that were or were not followed up (Table 3). This association also was maintained in the subgroup of 134 subjects that continued to periodically attend our outpatient clinic for several years after undergoing liver biopsy (Table 4). During the follow-up period, 9 of the 134 subjects developed HCC (all males, with a mean age of 55 ± 9 yrs; 4 patients were teetotalers, 5 reported drinking < 25 g/day of alcohol; 8 patients were HCV positive, and 1 patient had cryptogenic liver disease). At the time of liver biopsy, seven of the nine patients had had a histologic diagnosis of severe chronic hepatitis and two had received a diagnosis of cirrhosis. None had hepatocellular dysplasia or other histologic features predicting the subsequent development of liver carcinoma that occurred between 24–115 months (median, 87.5 mos) after liver biopsy. Six of the eight HCV-positive individuals had been treated with interferon but none was a long-term responder to the therapy.

Table 3. Demographic Characteristics and Histology of the Study Population According to the Occult HBV Status
 Occult HBV-positive cases (n = 135)Occult HBV-negative cases (n = 245)P
  1. HBV: hepatitis B virus; SD: standard deviation; NS: not significant.

Mean age in yrs (± SD)48 ± 1349 ± 12.8NS
Male/female ratio92/43154/91NS
Histologic diagnosis   
 Minimal changes1350 
 Chronic hepatitis921590.01
Follow-up patients5381NS
Table 4. Demographic Characteristics, Histology, HCV-Genotype Distribution, and HCC Development in 134 Follow-Up Patients According to their Occult HBV Status
 Occult HBV-positive (n = 53)Occult HBV-negative (n = 81)P
  1. HCV: hepatitis C virus; HCC: hepatocellular carcinoma; HBV: hepatitis B virus; SD: standard deviation; NS: not significant.

Mean age in yrs (± SD)51 ± 10.4552 ± 10.6NS
Male/female ratio34/1951/30NS
Mean follow-up in mos (± SD)76.8 ± 29.485.92 ± 33.48NS
HCV-positive patients5074NS
HCV-negative patients37NS
Total no. of available HCV genotypes4769 
No. of 1/2/3 genotypes30/13/444/17/8NS
Histologic diagnosis   
 Minimal changes314 
 Chronic hepatitis36600.007
HCC (no. of patients)810.002

Eight of these nine patients (seven of whom were HCV positive, and one of whom had cryptogenic liver disease) were positive for occult HBV and one patient (who was HCV positive) had no evidence of occult HBV infection (P = 0.002). It is interesting to note that six of the nine cases were anti-HBc positive (all of whom were occult HBV positive), whereas three patients (two who were occult HBV positive and one who was negative for occult HBV) were negative for all serum HBV markers (data not shown).

The diagnosis of HCC was performed through imaging techniques and histologically confirmed in all cases.


HBV is a major causative agent of liver carcinoma worldwide, and therefore has been classified as a Group 1 human carcinogen and is considered the second most important oncogenic agent after smoking tobacco.22, 23 There is much evidence suggesting that HBV is a risk factor for the development of HCC in the case of occult infection as well, and this occurs both in patients with HCV-related and cryptogenic chronic liver disease.1–3, 24 Moreover, we recently demonstrated that the potential mechanisms whereby overt HBV might induce tumor formation are maintained for the most part in the occult status.13 However, all the studies performed in the field were cross-sectional examinations of cases in which liver carcinoma already occurred, whereas to our knowledge no prospective study to date has investigated the fate of chronic hepatitis patients in terms of the development of HCC in accordance with their occult HBV status.

To our knowledge, the current observational cohort study is the first to evaluate the occurrence of HCC in a large number of HBsAg-negative patients with chronic liver disease who had been tested for occult HBV and then followed for more than 4 years. It is important to note that the presence of HBV genomes was investigated through the analysis of liver biopsy specimens, which currently is considered the gold standard for the detection of occult HBV.2, 3, 25 We found that this peculiar form of infection was associated (with a high statistical significance) with the development over time of liver carcinoma, because eight of the nine patients who developed HCC were positive for HBV DNA. It is interesting to note that the histologic examination of the liver biopsy specimens in no case had demonstrated any pattern predicting the future neoplastic transformation. In addition, two of the occult-infected subjects were negative for all HBV serum markers, confirming that occult HBV also may be present in individuals without any serologic evidence of previous infection and that the molecular biology approach for detecting viral DNA is the most sensitive and specific method for identifying the hidden virus.

We believe the results of the current study provide further evidence that occult HBV may be associated with the neoplastic transformation of a chronic liver disease, and definitively confirm that the presence of HBV genomic sequences distinguishes those individuals at higher risk of developing HCC among HBsAg-negative individuals. Consequently, testing chronic hepatitis patients for HBV genomes appears to be a powerful tool for the identification of subjects who need to be more carefully monitored to allow early diagnosis of HCC.