Complicated relationships of amino acid substitution in hepatitis C virus core region and IL28B genotype influencing hepatocarcinogenesis

Authors


  • Potential conflict of interest: Norio Akuta has received speakers' bureau from MSD K.K., and holds a right to get some loyalty from SRL. Inc.. Hiromitsu Kumada has received speakers' bureau from MSD K.K., Mitsubishi Tanabe Pharma, Dainippon Sumitomo Pharma, Bristol-Myers Squibb, and holds a right to get some loyalty from SRL. Inc.. Fumitaka Suzuki has received speakers' bureau from Bristol-Myers Squibb. The other authors have nothing to disclose.

  • Supported in part by Grants-in-Aid for scientific research and development from the Ministry of Health, Labor and Welfare and Ministry of Education Culture Sports Science and Technology, Government of Japan.

Abstract

The impact of amino acid (aa) 70 substitution in the core region on hepatocarcinogenesis and survival for liver-related death in patients of hepatitis C virus (HCV) genotype 1b (HCV-1b), who had not received antiviral therapy, is unknown. The relationships among aa 70 substitution, IL28B genotype, and hepatocarcinogenesis are also not clear. A total of 1,181 consecutive HCV-infected patients, who had not received antiviral therapy, were included in a follow-up study to determine predictive factors of hepatocarcinogenesis and survival for liver-related death. The cumulative hepatocarcinogenesis rates in HCV-1b of Gln70(His70) (glutamine (histidine) at aa 70) were significantly higher than those in HCV-1b of Arg70 (arginine at aa 70) and HCV-2a/2b. The cumulative survival rates for liver-related death in HCV-1b of Gln70(His70) were significantly lower than those in HCV-1b of Arg70 and HCV-2a/2b. Multivariate analysis identified gender (male), age (≥60 years), albumin (<3.9 g/dL), platelet count (<15.0 × 104/mm3), aspartate aminotransferase (≥67 IU/L), and HCV subgroup (HCV-1b of Gln70(His70)) as determinants of both hepatocarcinogenesis and survival rates for liver-related death. In HCV-1b patients, the cumulative change rates from Arg70 to Gln70(His70) by direct sequencing were significantly higher than those from Gln70(His70) to Arg70. In patients of Arg70 at the initial visit, the cumulative change rates from Arg70 to Gln70(His70) in IL28B rs8099917 non-TT genotype were significantly higher than those in the TT genotype. Conclusion: Substitution of aa 70 in the core region of HCV-1b is an important predictor of hepatocarcinogenesis and survival for liver-related death in HCV patients who had not received antiviral therapy. The IL28B genotype might partly affect changes over time of dominant amino acid in core aa 70 of HCV-1b. (HEPATOLOGY 2012;56:2134–2141)

Hepatitis C virus (HCV) usually causes chronic infection that can result in chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma (HCC).1, 2 At present, treatments based on interferon (IFN), in combination with ribavirin, are the mainstay for combating HCV infection. In Japan, HCV genotype 1b (HCV-1b) and high viral loads account for more than 70% of HCV infections, making it difficult to treat patients with chronic hepatitis C.3

Despite numerous lines of epidemiologic evidence connecting HCV infection and the development of HCC, it remains controversial whether HCV itself plays a direct role or an indirect role in the pathogenesis of HCC.4 It has become evident that HCV core region has oncogenic potential through the use of transgenic mice, but the clinical impact of the core region on hepatocarcinogenesis is still unclear.5 Previous reports indicated that amino acid (aa) substitutions at position 70 in the HCV core region of patients infected with HCV-1b are pretreatment predictors of poor virological response to pegylated IFN (PEG-IFN)/ribavirin combination therapy and triple therapy of telaprevir/PEG-IFN/ribavirin,6-9 and also affects hepatocarcinogenesis.10-13 These reports support the findings of oncogenic potential by core region from the clinical aspect. However, its impact on hepatocarcinogenesis and survival for liver-related death in patients of HCV-1b who had not received antiviral therapy is still unknown.

The IL28B genotype is a poor predictor of virological response to PEG-IFN/ribavirin combination therapy and triple therapy of telaprevir/PEG-IFN/ribavirin9, 14-17 and is reported to be associated with HCC, although its impact on HCC is controversial.18-21 Furthermore, treatment-resistant substitution of core aa 70 (glutamine/histidine at aa 70 (Gln70/His70)), which might affect hepatocarcinogenesis, was significantly more frequent in patients with treatment-resistant genotype (non-TT) than -sensitive genotype (TT) at IL28B rs8099917.21-23 Thus, the significant linkage between substitution of aa 70 and IL28B genotype had been shown, but it is not clarified whether the existence of a complex interaction between the virus and host might affect hepatocarcinogenesis.

The present study included 1,181 consecutive HCV-infected patients who had not received antiviral therapy. The aims of the study were: (1) To evaluate the impact of aa substitutions in the core region of HCV-1b on hepatocarcinogenesis and survival for liver-related death; and (2) To investigate the association of IL28B genotype and time-dependent aa changes in the core region of HCV-1b.

Patients and Methods

Patients.

Among 2,799 consecutive HCV-infected patients in whom antiviral therapy (IFN monotherapy or IFN plus ribavirin combination therapy) was not induced between December 1962 and November 2010 at Toranomon Hospital, 1,181 were selected in the present study based on the following criteria. (1) Positive for anti-HCV (third-generation enzyme immunoassay, Chiron, Emerville, CA) and positive for HCV RNA (nested polymerase chain reaction [PCR]), at the initial visit. (2) Patients without HCC at the initial visit. (3) Patients infected with single genotype of HCV-1b, 2a, or 2b. (4) In HCV-1b, patients analyzed aa substitutions of the core region by direct sequencing, one or more times from the initial visit. (5) Patients negative for hepatitis B surface antigen (radioimmunoassay, Dainabot, Tokyo, Japan). (6) Patients free of coinfection with human immunodeficiency virus. (7) Patients free of other types of chronic liver disease, including hemochromatosis, Wilson's disease, primary biliary cirrhosis, alcoholic liver disease, autoimmune liver disease, inherited liver disease including alpha-1 antitrypsin deficiency, and hepatic venous outflow block. (8) Patients who consented to the study.

Table 1 summarizes the profiles and laboratory data at the initial visit of 1,181 patients infected with HCV who had not received antiviral therapy. They did not receive antiviral therapy based on concerns about adverse effects, lack of time for treatment, physician recommendation based on the appearance of depression and cardiopulmonary disease, lower levels of aspartate aminotransferase (AST) / alanine aminotransferase (ALT), or elderly patients. They included 608 males and 573 females, aged 20 to 93 years (median, 60 years). The median follow-up time from the initial visit until death or until the last visit was 9.0 years (range, 0.0-37.7 years). The study protocol was approved by the Human Ethics Review Committee of the institution.

Table 1. Profiles and laboratory data at the initial visit of 1,181 patients infected with HCV, who had not received antiviral therapy
  1. Data are number and percentages of patients, except those denoted by *, which represent the median (range) values.

Demographic data 
 Number of patients1,181
 Sex (male/female)608/573
 Age (years)*60 (20-93)
 History of blood transfusion526 (49.2%)
 Family history of liver disease201 (20.3%)
 Lifetime cumulative alcohol intake (>500 kg)110 (10.8%)
Laboratory data* 
 Total bilirubin (mg/dl)0.7(0.1-20.0)
 Aspartate aminotransferase (IU/l)71 (13-1,052)
 Alanine aminotransferase (IU/l)88 (4-1,210)
 Albumin (g/dl)4.1 (1.O-5.5)
 Hemoglobin (g/dl)14.0 (7.8-18.0)
 Platelet count (× 104/mm3)15.3 (2.6-52.9)
 HCV genotype (lb / 2a or 2b)750/431
 Levels of viremia (high viral load)757 (74.4%)
Amino acid substitutions in the HCV genotype lb 
 Core aa 70 (argjnine / glutamine (histidine))431/319
 Core aa 91 (leucine / methionine)482/268

Laboratory Investigations.

Blood samples were frozen at −80°C within 4 hours of collection and were not thawed until used for testing. Anti-HCV, HCV RNA, HCV genotype, and aa substitutions of the HCV-1b core region were assayed using stored frozen sera. HCV genotype was determined by PCR using a mixed primer set derived from nucleotide sequences of the NS5 region.24 HCV RNA quantitative analysis was measured by branched DNA assay v. 2.0 (Chiron), AMPLICOR GT HCV Monitor v. 2.0 using the 10-fold dilution method (Roche Molecular Systems, Pleasanton, CA), or COBAS TaqMan HCV test (Roche Diagnostics, Tokyo, Japan). High viral load of viremia levels was defined as branched DNA assay ≥1.0 Meq/mL, AMPLICOR GT HCV Monitor ≥100 × 103 IU/mL, or COBAS TaqMan HCV test ≥5.0 log IU/mL. Low viral load was defined as branched DNA assay <1.0 Meq/mL, AMPLICOR GT HCV Monitor <100 × 103 IU/mL, or COBAS TaqMan HCV test <5.0 log IU/mL.

Detection of Amino Acid Substitutions in Core Regions of HCV-1b.

In the present study, aa substitutions of the core region of HCV-1b were analyzed by direct sequencing. HCV RNA was extracted from serum samples and reverse transcribed with random primer and MMLV reverse transcriptase (Takara Syuzo, Tokyo, Japan). Nucleic acids of the core region were amplified by nested PCR using the following primers. The first-round PCR was performed with CE1 (sense, 5′-GTC TGC GGA ACC GGT GAG TA-3′, nucleotides: 134-153) and CE2 (antisense, 5′-GAC GTG GCG TCG TAT TGT CG-3′, nucleotides: 1096-1115) primers, and the second-round PCR with CC9 (sense, 5′-ACT GCT AGC CGA GTA GTG TT-3′, nucleotides: 234-253) and CE6 (antisense, 5′-GGA GCA GTC GTT CGT GAC AT-3′, nucleotides: 934-953) primers. All samples were initially denatured at 95°C for 2 minutes. The 35 cycles of amplification were set as follows: denaturation for 30 seconds at 95°C, annealing of primers for 30 seconds at 55°C, and extension for 1 minute at 72°C with an additional 7 minutes for extension. Then 1 μL of the first PCR product was transferred to the second PCR reaction. Other conditions for the second PCR were the same as the first PCR, except that the second PCR primers were used instead of the first PCR primers. The amplified PCR products were purified by the QIA quick PCR purification kit (Qiagen, Tokyo, Japan) after agarose gel electrophoresis and then used for direct sequencing. Dideoxynucleotide termination sequencing was performed with the Big Dye Deoxy Terminator Cycle Sequencing kit (Perkin-Elmer, Tokyo, Japan).

With the use of HCV-J (accession no. D90208) as a reference,25 the dominant sequence of 1-191 aa in the core protein of HCV-1b was determined by direct sequencing and then compared with the consensus sequence constructed on 50 clinical samples to detect substitutions at aa 70 of arginine (Arg70) or glutamine/histidine (Gln70/His70) and aa 91 of leucine (Leu91) or methionine (Met91).6 Especially, patients were classified into three HCV subgroups according to HCV genotype in combination with aa substitutions in HCV-1b core region (HCV-1b of Arg70, HCV-1b of Gln70(His70), and HCV-2a/2b).

Determination of IL28B Genotype.

IL28B rs8099917 was genotyped by the Invader assay, TaqMan assay, or direct sequencing, as described.26, 27

Follow-Up and Diagnosis of HCC.

Follow-up of patients was made on a monthly to trimonthly basis after the initial visit. Imaging diagnosis was made one or more times per year with ultrasonography, computed tomography, or magnetic resonance imaging. During this time, liver-related death, which included HCC, cholangiocellular carcinoma, liver failure, or esophageal variceal bleeding, was also evaluated.

Statistical Analysis.

The cumulative rates of hepatocarcinogenesis, survival for liver-related death, and amino acid changes in the core region were calculated using the Kaplan-Meier technique; differences between the curves were tested using the log-rank test. Statistical analyses of hepatocarcinogenesis, survival, and amino acid changes, according to groups, were calculated using the period from the initial visit. Stepwise Cox regression analysis was used to determine independent predictive factors that were associated with hepatocarcinogenesis and survival for liver-related death. The hazard ratio (HR) and 95% confidence interval (95% CI) was also calculated. Potential predictive factors associated with hepatocarcinogenesis and survival for liver-related death included the variables: sex, age, history of blood transfusion, family history of liver disease, lifetime cumulative alcohol intake, total bilirubin, AST, ALT, albumin, hemoglobin, platelet count, levels of viremia, and HCV subgroup according to HCV genotype in combination with aa substitution in core region. Variables that achieved statistical significance (P < 0.05) on univariate analysis were tested by multivariate Cox proportional hazard model to identify significant independent factors. Statistical comparisons were performed using the SPSS software (Chicago, IL). P < 0.05 by the two-tailed test were considered significant.

Abbreviations

aa, amino acid; ALT, alanine aminotransferase; AST, aspartate aminotransferase; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; PEG/IFN, pegylated interferon.

Results

Hepatocarcinogenesis Rates and Survival Rates for Liver-Related Death in Patients Infected With HCV Who Had Not Received Antiviral Therapy.

During the follow-up, 413 patients (35.0%) developed HCC. The cumulative hepatocarcinogenesis rates were 16.3, 34.3, 48.3, 58.7, and 69.1% at the end of 5, 10, 15, 20, and 25 years, respectively. The median interval between the initial visit and detection of HCC was 6.2 years (range, 0.1-31.7 years).

During the follow-up period, 243 patients (20.6%) died due to liver-related causes, and 97 of 243 (90.5%) developed HCC. The cumulative survival rates for liver-related death were 96.2, 84.8, 68.9, 55.0, and 46.1% at the end of 5, 10, 15, 20, and 25 years, respectively. The median interval between the initial visit and liver-related death was 10.1 years (range, 0.4-35.8 years).

Hepatocarcinogenesis Rates and Survival Rates for Liver-Related Death According to HCV Genotype in Combination with Amino Acid Substitutions in Core Region of HCV-1b.

During the follow-up, 163 patients (51.3%), 175 (41.2%), and 75 (17.6%) developed HCC in HCV-1b of Gln70(His70), HCV-1b of Arg70, and HCV-2a/2b, respectively. In HCV-1b of Gln70(His70), HCV-1b of Arg70, and HCV-2a/2b, the cumulative hepatocarcinogenesis rates were 21.7, 19.3, 8.0% at the end of 5 years; 44.4, 39.4, 18.2% at the end of 10 years; 60.4, 52.7, 29.1% at the end of 15 years; 71.6, 60.3, 43.1% at the end of 20 years; and 87.1, 69.8, 46.9% at the end of 25 years, respectively. The rates were significantly different among the three HCV subgroups (P < 0.001) (Fig. 1). Especially, the rates in HCV-1b of Gln70(His70) were significantly higher than those in HCV-1b of Arg70 (P = 0.028) and HCV-2a/2b (P < 0.001), and the rates in HCV-1b of Arg70 were also significantly higher than those in HCV-2a/2b (P < 0.001).

Figure 1.

Cumulative hepatocarcinogenesis rates according to HCV genotype in combination with amino acid substitutions in core region of HCV-1b. The rates were significantly different among the three HCV subgroups (P < 0.001; log-rank test). Especially, the rates in HCV-1b of Gln70(His70) were significantly higher than those in HCV-1b of Arg70 (P = 0.028; log-rank test) and HCV-2a/2b (P < 0.001; log-rank test), and the rates in HCV-1b of Arg70 were also significantly higher than those in HCV-2a/2b (P < 0.001; log-rank test).

During the follow-up, 104 patients (34.4%), 97 (23.4%), and 42 (10.0%) died due to liver-related causes in HCV-1b of Gln70(His70), HCV-1b of Arg70, and HCV-2a/2b, respectively. In HCV-1b of Gln70(His70), HCV-1b of Arg70, and HCV-2a/2b, the cumulative survival rates for liver-related death were 95.2, 95.4, 97.9% at the end of 5 years; 77.7, 83.3, 93.9% at the end of 10 years; 58.4, 68.4, 81.2% at the end of 15 years; 39.3, 58.4, 69.0% at the end of 20 years; and 33.8, 47.5, 59.5% at the end of 25 years, respectively. The rates were significantly different among the three HCV subgroups (P < 0.001) (Fig. 2). Especially, the rates in HCV-1b of Gln70(His70) were significantly lower than those in HCV-1b of Arg70 (P = 0.016) and HCV-2a/2b (P < 0.001), and the rates in HCV-1b of Arg70 were also significantly lower than those in HCV-2a/2b (P < 0.001).

Figure 2.

Cumulative survival rates for liver-related death according to HCV genotype in combination with amino acid substitutions in the core region of HCV-1b. The rates were significantly different among the three HCV subgroups (P < 0.001; log-rank test). Especially, the rates in HCV-1b of Gln70(His70) were significantly lower than those in HCV-1b of Arg70 (P = 0.016; log-rank test) and HCV-2a/2b (P < 0.001; log-rank test), and the rates in HCV-1b of Arg70 were also significantly lower than those in HCV-2a/2b (P < 0.001; log-rank test).

Predictive Factors Associated with Hepatocarcinogenesis and Survival for Liver-Related Death in Patients Infected with HCV Who Had Not Received Antiviral Therapy.

The data for the whole population sample were analyzed to determine those factors that could predict hepatocarcinogenesis and survival for liver-related death.

Univariate analysis identified eight parameters that significantly correlated with hepatocarcinogenesis. These included gender (male; P < 0.001), age (≥60 years; P < 0.001), total bilirubin (≥1.2 mg/dL; P < 0.001), AST (≥67 IU/L; P < 0.001), ALT (≥85 IU/L; P < 0.001), platelet count (<15.0 × 104/mm3; P < 0.001), albumin (<3.9 g/dL; P < 0.001), and lifetime cumulative alcohol intake (≥500 kg; P = 0.025). Furthermore, the rates in HCV-1b of Gln70(His70) were significantly higher than those in HCV-1b of Arg70 (P = 0.028) and HCV-2a/2b (P < 0.001). These factors were entered into multivariate analysis, which then identified six parameters that significantly influenced hepatocarcinogenesis independently: gender (male; HR 1.78, P < 0.001), age (≥60 years; HR 1.68, P < 0.001), albumin (<3.9 g/dL; HR 1.94, P < 0.001), platelet count (<15.0 × 104/mm3; HR 2.89, P < 0.001), AST (≥67 IU/L; HR 1.92, P < 0.001), and HCV subgroup (HCV-1b of Gln70(His70); HR 1.94, P = 0.001) (Table 2).

Table 2. Factors associated with hepatocarcinogenesis in patients infected with HCV, who had not received antiviral therapy, identified bymultivariate analysis
[Factors][Category]Hazard ratio (95% confidence interval)P
  1. Cox proportional hazard model

Gender1: female1 
 2: male1.78 (1.44-2.21)<0.001
Age (years)l:<601 
 2:≥601.68 (1.35-2.09)<0.001
Albumin (g/dl)1: ≥3.91 
 2: <3.91.94 (1.55-2.42)<0.001
Platelet count (× 104/mm3)1: ≥15.01 
 2:<15.02.89 (2.25-3.72)<0.001
Aspartate aminotransferase (IU/l)1:<671 
 2:≥671.92 (1.47-2.52)<0.001
HCV subgroup1: HCV-2a/2b1 
 2: HCV-1b with Arg701.91 (1.42-2.55)<0.001
 3: HCV-1b with Gln70(His70)1.94 (1.45-2.61)<0.001

Univariate analysis identified seven parameters that significantly correlated with survival for liver-related death. These included gender (male; P < 0.001), age (≥60 years; P < 0.001), total bilirubin (≥1.2 mg/dL; P < 0.001), AST (≥67 IU/L; P < 0.001), ALT (≥85 IU/L; P < 0.001), platelet count (<15.0 × 104/mm3; P < 0.001), and albumin (<3.9 g/dL; P < 0.001). Furthermore, the rates in HCV-1b of Gln70(His70) were significantly lower than those in HCV-1b of Arg70 (P = 0.016) and HCV-2a/2b (P < 0.001). These factors were entered into multivariate analysis, which then identified six parameters that significantly influenced survival for liver-related death independently: gender (male; HR 1.91, P < 0.001), age (≥60 years; HR 1.61, P = 0.001), albumin (<3.9 g/dL; HR 2.49, P < 0.001), platelet count (<15.0 × 104/mm3; HR 3.69, P < 0.001), AST (≥67 IU/L; HR 4.16, P < 0.001), and HCV subgroup (HCV-1b of Gln70(His70); HR 2.16, P < 0.001) (Table 3).

Table 3. Factors associated with survival for liver-related death in patients infected with HCV, who had not received antiviral therapy, identified by multivariate analysis
[Factors][Category]Hazard ratio (95% confidence interval)P
  1. Cox proportional hazard model

Gender1: female1 
 2: male1.91 (1.45-2.52)<0.001
Age (years)l:<601 
 2:≥601.61 (1.21-2.12)0.001
Albumin (g/dl)1:≥3,91 
 2:<3.92.49 (1.87-3.31)<0.001
Platelet count (× 104/mm3)1:≥15.01 
 2:<15.03.69 (2.65-5.13)<0.001
Aspartate aminotransferase (IU/l)1:<671 
 2:≥674.16 (2.43-7.11)<0.001
HCV subgroup1: HCV-2a/2b1 
 2: HCV-1b with Arg701.83 (1.25-2.68)0.002
 3: HCV-1b with Gln70(His70)2.16 (1.48-3.16)<0.001

IL28B Genotype and Time-Dependent Amino Acid Changes in Core Region of HCV-1b.

Among 1,181 patients, 359 could be evaluated for changes over time of dominant amino acid by direct sequencing in core aa 70 of HCV-1b. Furthermore, among 359 patients, 142 could also be analyzed for the relationship between IL28B rs8099917 genotype and time-dependent changes of core aa 70.

In 199 patients of Arg70 at the initial visit, 34 patients (17.1%) changed from Arg70 to Gln70(His70) during the follow-up. Inversely, in 160 patients of Gln70(His70) at the initial visit, eight patients (5.0%) changed from Gln70(His70) to Arg70 during the follow-up. In change from Arg70 to Gln70(His70), and change from Gln70(His70) to Arg70, the cumulative change rates were 3.0, 0% at the end of 5 years; 16.8, 5.8% at the end of 10 years; 27.4, 11.5% at the end of 15 years; and 38.9, 16.7% at the end of 20 years, respectively. The cumulative change rates from Arg70 to Gln70(His70) were significantly higher than those from Gln70(His70) to Arg70 (P = 0.002).

In 78 patients of Arg70 and TT genotype at the initial visit, nine (11.5%) changed from Arg70 to Gln70(His70) during the follow-up. In 11 patients of Arg70 and non-TT genotype at the initial visit, seven (63.6%) changed from Arg70 to Gln70(His70) during the follow-up. In TT and non-TT genotype, the cumulative change rates were 0, 9.1% at the end of 5 years; 3.2, 65.4% at the end of 10 years; 14.8, 65.4% at the end of 15 years; and 29.0, 65.4% at the end of 20 years, respectively. The cumulative change rates in non-TT genotype were significantly higher than those in TT genotype (P < 0.001) (Fig. 3A).

Figure 3.

Changes over time of dominant amino acid by direct sequencing in core aa 70 of HCV-1b, according to IL28B rs8099917 genotype. (A) In HCV-1b patients of Arg70 at the initial visit, cumulative change rates from Arg70 to Gln70(His70) during follow-up. The rates in non-TT genotype were significantly higher than those in TT genotype (P < 0.001; log-rank test). (B) In HCV-1b patients of Gln70(His70) at the initial visit, cumulative change rates from Gln70(His70) to Arg70 during follow-up. The rates in TT genotype were not significantly higher than those in non-TT genotype (P = 0.114; log-rank test).

In 30 patients of Gln70(His70) and TT genotype at the initial visit, three patients (10.0%) changed from Gln70(His70) to Arg70 during the follow-up. In 23 patients of Gln70(His70) and non-TT genotype at the initial visit, no patients changed from Gln70(His70) to Arg70 during the follow-up. In TT and non-TT genotype, the cumulative change rates were 0, 0% at the end of 5 years; 9.1, 0% at the end of 10 years; 20.5, 0% at the end of 15 years; and 20.5, 0% at the end of 20 years, respectively. The cumulative change rates in TT genotype were not significantly higher than those in non-TT genotype (P = 0.114) (Fig. 3B).

Discussion

This is the first report to indicate that aa substitution in the core region might affect hepatocarcinogenesis and survival for liver-related death in HCV patients who had not received antiviral therapy. The treatment-resistant mechanism and oncogenic potential of HCV core region are still unclear. Moriishi et al.28, 29 showed that a knockout of the PA28γ gene induces the accumulation of HCV core protein in the nucleus of hepatocytes of HCV core gene transgenic mice and disrupts development of both hepatic steatosis and HCC. Hu et al.13 indicated that the point-mutations of the core gene, including core aa 70 and aa 91, might change the secondary structure of not only RNA but also protein. As a result, the functions of both RNA and protein of the core region, such as an interaction with other DNA/RNA or proteins, might change and lead to hepatocarcinogenesis. Funaoka et al.30 recently reported that treatment-resistant substitutions of core aa 70 and aa 91 (Gln70/His70 and Met91) were resistant to interferon in vitro, and the resistance might be induced by interleukin 6-induced upregulation of SOCS3. Further studies should be performed to investigate the treatment-resistant mechanism and oncogenic potential of aa substitution in the core region.

The association between HCV genotype and the risk of HCC is not clear. A previous report indicated that hepatocarcinogenesis rates in patients infected with HCV-1b were significantly higher than those in patients infected with HCV-2a/2c, based on an Italian cohort,31 and this finding might be partly explained by distribution of HCV-1b of Arg70 or Gln70(His70). In fact, the hepatocarcinogenesis rates in HCV-1b of Gln70(His70) were significantly higher than those in HCV-1b of Arg70 and HCV-2a/2b in the present study based on a Japanese cohort. The present study is the first report to indicate that substitution of aa 70 in the core region of HCV-1b is not only an important predictor of hepatocarcinogenesis, but also of survival for liver-related death in HCV patients who had not received antiviral therapy. The reason for the higher rates of liver-related death in HCV-1b of Gln70(His70) might be due to the higher rates of HCC. In conclusion, reducing the risk of hepatocarcinogenesis by HCV RNA eradication and/or ALT normalization by antiviral therapy should be recommended, especially in HCV-1b of Gln70(His70) as a high-risk group for hepatocarcinogenesis.32

The significant linkage between substitution of aa 70 and IL28B genotype had been shown,21-23 but the mechanism of complex interaction between the virus and host is not clear. In the present study, the cumulative change rates from Arg70 to Gln70(His70) were significantly higher than those from Gln70(His70) to Arg70. Especially, the rates from Arg70 to Gln70(His70) in IL28B rs8099917 non-TT genotype were significantly higher than those in TT genotype. Although the molecular mechanisms of their relationship remain unknown, it could be speculated that IL28B genotype has an influence on the time-dependent changes of core aa 70, and refractory factors for treatment might accumulate in HCV-1b patients with non-TT. Hence, elucidating the relationship between substitution of aa 70 and IL28B genotype is an important step in understanding the mechanism of HCV treatment-resistance and disease progression.

The impact of IL28B genotype on hepatocarcinogenesis is controversial.18-21 In this study, the effect of IL28B rs8099917 genotype on HCC was assessed in 515 of 2,799 consecutive HCV-infected patients who had not received antiviral therapy. Interestingly, the cumulative hepatocarcinogenesis rates in TT of the treatment-sensitive genotype was not significantly lower than those in non-TT of the treatment-resistant genotype (P = 0.930; log-rank test) in a preliminary study based on a small numbers of patients (Fig. 4). This result suggests that core aa 70 as a predictor of hepatocarcinogenesis might not only be influenced by IL28B genotype, but also by other factors strongly related to hepatocarcinogenesis independent of IL28B genotype. As a whole, it is regrettable that its impact on hepatocarcinogenesis in HCV patients who had not received antiviral therapy could not be investigated in this study. Further comprehensive studies should be performed to disclose the molecular mechanisms for the complicated relationships among core aa 70, IL28B genotype, and hepatocarcinogenesis.

Figure 4.

Cumulative hepatocarcinogenesis rates according to IL28B rs8099917 genotype. The rates in TT genotype were not significantly lower than those in non-TT genotype (P = 0.930; log-rank test) in a preliminary study based on a small number of 515 patients.

The limitations of the present study are that patients who had received treatment besides IFN-related therapy (such as ursodeoxycholic acid, branched chain amino acid, and phlebotomy) could not be excluded. Furthermore, the clinical impact of metabolic factors (such as diabetes, insulin resistance, hepatocyte steatosis, and obesity) on hepatocarcinogenesis could also not be investigated. Further studies should be performed to investigate the clinical impact of treatment besides IFN-related therapy and metabolic factors on hepatocarcinogenesis.33-37

In conclusion, substitution of aa 70 in the core region of HCV-1b is the important predictor of hepatocarcinogenesis and survival for liver-related death in HCV patients who had not received antiviral therapy. This study emphasizes the importance of antiviral therapy to reduce the risk of hepatocarcinogenesis, especially in HCV-1b of Gln70(His70) as a high-risk group for hepatocarcinogenesis. Furthermore, IL28B genotype might partly affect changes over time of dominant amino acid in core aa 70. This result should be interpreted with caution because races other than Japanese populations and patients infected with HCV-1a were not included. Any generalization of the results should await confirmation by studies of patients of other races and HCV-1a. Further prospective studies of a larger number of patients matched for race and HCV genotype are required to explore the relationship between core aa 70, IL28B genotype, and hepatocarcinogenesis.

Ancillary