Role of hepatitis B virus genotypes Ba and C, core promoter and precore mutations on hepatocellular carcinoma: a case control study

the clinical on presentation, the chance of receiving treatment and the cumulative survival rate for chemoembolization-treated patients between patients with subtype Ba and genotype C. There was too small a number of CP wild-type to do a similar comparison with CP mutants. In conclusion, there was a significantly higher prevalence of both genotype C and CP mutations in patients with HCC. The association between HBV genotype C and HCC was probably not genuine but was due to the high percentage of CP mutations in patients with genotype C.


Introduction
Hepatitis B virus (HBV) genotypes have been shown to influence the disease profile of chronic hepatitis B infection in some recent studies (1EE9). In Asia where HBV genotypes B and C are commonly found, the natural history of patients with genotype B differs from those with genotype C. The former had a higher rate and earlier age of hepatitis B e antigen (HBeAg) seroconversion, less serious liver diseases and better interferon treatment response compared with the latter (1EE7, 10,11).
However, the exact role of these two genotypes on the development of hepatocellular carcinoma (HCC) is still controversial. According to an early Taiwan study, patients with genotype B are associated with the development of HCC at young age (550 years old) (12). Subsequent studies conducted in Japan, however show contradictory findings. HBV genotype C is more commonly found in HCC patients below age 50 (13,14). According to Orito and his colleagues, the mean age of the HCC patients with genotype B is older than that of patients with genotype C (15). Another study conducted by Sumi and his colleagues report that although patients with genotype C compared with patients with genotype B are associated with a slower development of HCC, there is no difference in the life-time risk of development of HCC (2). The role of HBV genotypes in the development of HCC therefore remains uncertain.
Patients with core promoter (CP) mutants compared with those with wild-type have also been shown to have a higher risk of development of HCC (16EE18). It has been demonstrated further that HBV genotype C is associated with CP mutations (1). It is important to elucidate the interplay between HBV genotypes and CP mutations since both are associated with the development of HCC.
Concerning the clinical features of HCC, a previous study examining nine patients (four with genotype B, five with genotype C) with HCC shows that patients with genotype B have a better response to embolization therapy and better survival compared with patients with genotype C (19). Large-scale studies are required to determine whether there are any differences in the clinical presentation, chance of receiving treatment and outcome of the HCC between patients with genotypes B and C.
The aims of the present study are to elucidate the exact role of HBV genotypes, CP and precore mutations on the development of HCC and to determine whether there were any differences between HCC patients with genotypes B and C in term of the clinical features, the chance of receiving treatment and survival. alcoholic liver disease, four had cryptogenic cirrhosis}. The remaining 102 patients were positive for HBsAg and negative for anti-HDV. Of these, 90 patients in whom the HBV genotype results were available were recruited for analyses in the present study. The remaining 12 patients obtained no PCR amplification for HBV genotyping.

Materials and methods
Blood was taken for measurement of complete blood count, liver biochemistry, prothrombin time, AFP, HBV serological markers including HBsAg, HBeAg and antibody to HBeAg (anti-HBe) [measured by the micro-particle enzyme immunoassay (MEIA), Abbott Laboratories, Chicago, IL]. Serum HBV DNA levels were measured by Digene Hybrid Capture assay, Digene Corporation, Gaithersburg, MD (lower limit of detection 0.14 Â 10 6 copies/ml).
All patients were assessed for the feasibility of surgical resection. For patients in whom surgical resections were not possible, they were assessed for the possibility of transarterial chemoembolization (TACE) treatment, which was performed every 8EE12 weeks continuously. TACE would be terminated in patients with progressive disease after five treatment courses. Tumor responses were assessed by the difference in the largest dimensions of the index tumor at the first and the last sessions of TACE at the time of writing. If TACE was not feasible, patients were treated conservatively.
To examine the role of HBV genotypes B and C in the development of HCC, every patient with HCC was matched for gender, age and HBeAg status with two HBsAg-positive patients without HCC. Therefore, the total number of disease controls was 180. These control patients are consecutive patients with chronic hepatitis B infection without HCC, being followed up during the period of recruitment of the present study in our Hepatitis Clinic, Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Hong Kong.
HBV genotyping for the 90 patients with HCC and 180 patients without HCC (controls) was determined by an enzyme-linked immunosorbent assay (ELISA) using monoclonal antibodies against seven distinct epitopes (b, m, k, s, u, f and g) on the preS2-region products of HBsAg (20). HBV genotype B subtypes were determined by a PCR-restriction fragment length polymorphism method as described in our previous study (21).
The CP and precore mutations were determined by direct sequencing. Briefly, serum HBV DNA, in a final volume of 200 ml, was extracted from 200 ml of serum, using the QIAGEN DNA blood minikit (QIAGEN GmbH, Germany). The HBV genome at nucleotides (nt) 1653EE1974 containing precore and CP regions was amplified by PCR using the primers HBV1 (5 H -CATAAGAGGACTCTTGGACT-3 H ) and HBV2 (5 H -GGAAAGAAGTCA-GAGGC-3 H ) in a GeneAmp 9700 PCR system (Applied Biosystems, US) with the following conditions: 95 C for 10 min, followed by 94 C for 30 s, 54 C for 30 s and 72 C for 1 min up to 40 cycles and a final extension 72 C for 10 min. The 322 bp PCR products were purified by ethanol precipitation with ammonium acetate in a final volume of 20 ml. Two microliters of the purified PCR products were used in a thermocycle sequencing with the DYEnamic ET Terminator Cycle Sequencing Kit as directed (Amersham Biosciences, Uppsala, Sweden) using HBV2 as the primer. Automated sequencing was performed with an ABI prism 3700 DNA Analyzer. For CP region, the wildtype (nt 1762A/1764G) and the mutants (nt 1762A/1764A, nt 1762A/1764T, nt 1762T/1764A) were determined. For the precore region, the wild-type (nt 1896G) and the mutant (nt 1896A) were determined.

Statistical analysis
All statistical analyses were performed using the Statistical Program for Social Sciences (SPSS 10.0 for windows, SPSS, Chicago, IL). Continuous variables with skewed distribution were tested by MannEEWhitney U test. Categorical variables were tested by x 2 test with Yates' correction or Fisher's exact test. Logistic regression analysis was used to test whether factors were independently associated with the development of HCC. The cumulative survival was examined by KaplanEEMeier analysis using log rank test. P values of 50.05 were considered as statistically significant.

Demographics
The demographic data of the 90 patients with HCC are listed in Table I. CP and precore mutations and HCC Sixty-six HCC patients and 135 control patients were available for analyses of CP and precore mutations (the unavailability of the remaining samples were due to the absence of the amplified DNA products after PCR). All the CP mutations identified in the HCC patients and control patients had double mutations, i.e. nt 1762T/1764A. All the precore mutations identified in HCC patients and control patients were nt 1896A. There was a significantly higher percentage of patients with CP mutations in the HCC group (60 out of 66, 90.9%) compared with that of the control group (101 out of 135, 74.8%; P 0.007, OR 3.4, 95% CI 1.3EE8.5). There was no difference in the percentage of

Clinical features of HCC patients with genotypes Ba and C
Since only six out of 66 HCC patients had CP wild-type, comparison of the clinical features between patients with CP mutants and those with wild-type could not be performed due to type II errors. The various comparisons between patients with subtype Ba and those with genotype C were analyzed below.

Liver function tests, HBV DNA levels and tumor characteristics
The demographic data, prevalence of CP and precore mutations, liver biochemistry, prothrombin time, platelet counts and HBV DNA levels of HCC patients with genotypes Ba and C are listed in Table II. There were no differences in any of the parameters. The characteristics of HCC in patients with genotypes Ba and C are also listed in Table II. There were also no differences in any tumor parameters between the two groups of patients. For TACE treatment, tumor responses defined as a 50% reduction of the largest dimension were observed in three patients (30%) with subtype Ba and 12 patients (38.7%, P 1.0) with genotype C. There was no difference in the percentage of patients with a reduction of AFP level, which was defined as a 50% reduction for those with an elevated AFP level at baseline, between patients with subtype Ba and those with genotype C [1/5 (20%) versus 4/23 (17.4%), respectively, P 1.0].

Survival
The cumulative survival rate was examined in 46 patients who received TACE (10 with subtype Ba, 36 with genotype C) ( Figure 2). There was no difference in the cumulative survival rate between patients with subtype Ba and genotype C (P 0.65). The median survival time was 45 and 47.5 months for patients with genotypes Ba and C, respectively. The number of patients receiving surgical resection or conservative treatment was too small for meaningful statistical analysis.

Discussion
All the patients with genotype B in the present study from Hong Kong were of subtype Ba. This supports the phenomenon observed by Sugauchi et al. that HBV genotype Bj is exclusively found in Japan and Ba in the rest of the countries of Asia (21,22). Any differences in the disease profile of HCC ≤ Fig. 1. The prevalence of subtype Ba and genotype C in patients with HCC and control patients.

Hepatitis B genotypes and hepatocellular carcinoma
between patients with subtypes Bj and Ba would require comparative studies of patients from different countries in Asia.
In the present study, HBV genotype C and CP mutants had a higher chance of development of HCC when compared with HBV subtype Ba and CP wild-type, respectively. However, the only independent factor that was associated with HCC was CP mutations. HBV genotype C was not an independent factor. It has been shown that HBV genotypes are associated with the presence or absence of CP mutations (1,2,5,15). The apparent association between genotype C and HCC as demonstrated initially in the present study as well as other previous studies (13,14) might be in fact due to the close association between HBV genotype C and CP mutations, with CP mutations being the real risk factor for HCC. However, when genotype B (mainly subtype Bj) was compared with genotype C in a previous Japanese study, carriers of genotype C were independently associated with advanced liver diseases such as HCC (14). This might depend on the incidence of CP mutations among subtypes Bj, Ba and genotype C. A recent case-control study indicated that the double mutation in the CP occurred significantly less often in carriers of HBV subtype Bj than those of subtype Ba (15 versus 33%, P 5 0.05) or genotype C (15 versus 63%, P 5 0.001) (21). It is currently believed that subtype Ba is derived from the recombination of subtype Bj with genotype C in the precore and core regions (22), spanning nucleotide position from 1740EE1838 to 2443EE2485 (including the cytotoxic T-cell epitopes). Thus, subtype Ba is similar to genotype C in having a higher disease inducing capacity than subtype Bj. Additionally, even in HCC patients with subtype Ba, the prevalence of CP mutations was extraordinarily high when compared with that of controls  (23). Secondly, the transcription of the pre-genomic RNA may also be increased by the CP mutations through the removal of the nuclear receptor-binding site and at the same time creating a binding site for hepatocyte nuclear factor (24). This combination of changes increases the core RNA transcription (enhances core protein, DNA polymerase, pre-genomic RNA synthesis) but suppresses the precore RNA transcription (normally decreases the pregenomic RNA packaging) (25). Therefore, the end result will be an increase in viral replication. However, this enhanced viral replication has not been confirmed from other crosssectional studies (7,26,27). Longitudinal studies monitoring viral load continuously are required to demonstrate the effects of CP mutations on the viral replication in chronic hepatitis B infection, which is a longstanding disease. Furthermore, the CP mutations (T1762/A1764) may induce not only amino acid change in X protein but also an alteration of HBV gene expression. The alternation of X protein might play a role in hepatocarcinogenesis, because its coding sequence overlaps regions of crucial importance for viral replication such as enhancer II and the CP (18,28). Besides, CP mutations may give rise to HCC through other unknown pathways that also require further investigations. Many previous studies (1EE7) show that patients with genotype B, compared with patients with genotype C have less aggressive disease as reflected by the higher HBeAg seroconversion and better histology. Further studies are required to determine whether these favorable parameters observed in patients with genotype B are due to the lower chance of having CP mutations. But, as far as the risk for HCC is concerned, the present study showed that patients with subtype Ba with concomitant CP mutations had a higher risk of development of HCC compared with that of patients with subtype Ba without CP mutations (OR 6.3, 95% CI 1.3EE30.0). There was no difference in the risk for HCC between patients with subtype Ba and genotype C when the confounding factor, i.e. CP mutations, was taken into consideration.
In addition, the present study showed that there were no differences in the liver function, the tumor characteristics, chance of receiving treatment and survival between patients with subtype Ba and genotype C. It was unfortunate that similar comparisons between patients with CP mutants and wild-type could not be carried out because the overwhelming majority of HCC patients, 91.0%, had CP mutations.
In conclusion, there was a significantly higher prevalence of both genotype C and CP mutations in patients with HCC. The association between HBV genotype C and HCC was probably not genuine but was due to the high percentage of CP mutations. Concomitant CP mutations in patients with subtype Ba increased the risk of development of HCC. There were no differences in clinical features, chance of receiving treatment and survival between HCC patients with subtype Ba and genotype C.