fax: (81) 44-860-1623.
Amino acid substitutions in the hepatitis C virus core region are the important predictor of hepatocarcinogenesis†
Article first published online: 26 JUL 2007
Copyright © 2007 American Association for the Study of Liver Diseases
Volume 46, Issue 5, pages 1357–1364, November 2007
How to Cite
Akuta, N., Suzuki, F., Kawamura, Y., Yatsuji, H., Sezaki, H., Suzuki, Y., Hosaka, T., Kobayashi, M., Kobayashi, M., Arase, Y., Ikeda, K. and Kumada, H. (2007), Amino acid substitutions in the hepatitis C virus core region are the important predictor of hepatocarcinogenesis. Hepatology, 46: 1357–1364. doi: 10.1002/hep.21836
Potential conflict of interest: Nothing to report.
- Issue published online: 29 OCT 2007
- Article first published online: 26 JUL 2007
- Manuscript Accepted: 25 MAY 2007
- Manuscript Received: 21 MAR 2007
- Ministry of Health, Labor, and Welfare, Japan
We showed previously that amino acid (aa) substitutions in hepatitis C virus core region (HCV-CR) are negative predictors of virologic response to pegylated interferon (IFN) plus ribavirin therapy. HCV-CR induces hepatocellular carcinoma in transgenic mice, but the clinical impact is still unclear. To evaluate the impact of aa substitutions in HCV-CR on hepatocarcinogenesis, we performed a follow-up study on 313 noncirrhotic consecutive naïve patients infected with HCV genotype 1b who received IFN monotherapy. The median follow-up was 14.7 years. A sustained virologic response (SVR) after the first IFN was achieved by 65 patients (20.8%) (group A). Of 248 patients (79.2%) of non-SVR after first IFN, 112 (35.8%) did not receive additional IFN (group B), and the remaining 136 (43.5%) received multicourse IFN monotherapy (group C). As a whole, cumulative hepatocarcinogenesis rates in double wild-type (arginine at aa 70/leucine at aa 91) of HCV-CR were significantly lower than those in nondouble wild-type. Multivariate analyses identified 3 parameters (fibrosis stage 3, nondouble wild-type of HCV-CR, and group B) that tended to or significantly influenced hepatocarcinogenesis independently. With regard to hepatocarcinogenesis rates in group C according to HCV-CR and the mean alanine aminotransferase (ALT) during IFN-free period, significantly higher rates were noted in patients of nondouble wild-type with ALT levels of more than 1.5 times the upper limit of normal (25.7%) compared with the others (2.4%). Conclusion: Amino acid substitutions in the HCV-CR are the important predictor of hepatocarcinogenesis. In multicourse IFN therapy to nondouble wild-type, we emphasize the importance of reducing the risk of hepatocarcinogenesis by mean ALT during an IFN-free period below 1.5 times the upper limit of normal. (HEPATOLOGY 2007.)
Hepatitis C virus usually causes chronic infection, which can result in chronic hepatitis, cirrhosis, and hepatocellular carcinoma (HCC).1–5 In patients with chronic HCV, treatment with IFN can induce viral clearance and marked biochemical and histological improvement.6, 7
For chronic HCV infection, peginterferon (PEG-IFN) plus ribavirin (RBV) combination therapy is an expensive treatment modality that is accompanied by severe side effects and high sustained virological response (SVR). Patients who do not achieve SVR need to be identified before the start of combination therapy to avoid unnecessary side effects and high costs. Thus, safer IFN monotherapy should be considered to reduce the risk of hepatocarcinogenesis in patients unsuitable for PEG-IFN plus RBV therapy. We studied previous determinants of response to PEG-IFN plus RBV in patients with high titers of HCV genotype 1b (≥100 KIU/mL), which is dominant in Japan. Our results identified substitution of amino acids (aa) 70 and/or 91 in the HCV core region (HCV-CR) as an independent and significant negative predictor associated with virological response.8–10 Furthermore, we reported that multicourse IFN monotherapy reduces the risk of hepatocarcinogenesis and increases survival even if patients fail to achieve SVR after a single-course IFN, and that low ALT levels during an IFN-free period is associated with lower rates of hepatocarcinogenesis.11 Hence, multicourse IFN monotherapy might be expected to reduce the risk of hepatocarcinogenesis in patients who have negative predictors for PEG-IFN plus RBV.
Despite numerous lines of epidemiological evidence connecting HCV infection and the development of HCC, it remains controversial whether HCV itself plays a direct or indirect role in the pathogenesis of HCC.12 It has become evident that HCV-CR has oncogenic potential through the use of transgenic mice, but the clinical impact of HCV-CR on hepatocarcinogenesis is still unclear.13 Whether substitution of aa 70 and/or 91 in HCV-CR as a predictor of virological response for PEG-IFN plus RBV therapy also affects hepatocarcinogenesis awaits further investigation.
The present study included 313 consecutive naïve cases infected with HCV genotype 1b in whom 15 years had elapsed since induction of IFN monotherapy. The aims of the study were: (1) to evaluate the clinical impact of aa substitutions in the HCV-CR on hepatocarcinogenesis; (2) to analyze the predictive factors associated with hepatocarcinogenesis in patients who received IFN monotherapy; and (3) to evaluate the long-term efficacy of multicourse IFN monotherapy on hepatocarcinogenesis as examined through analysis of the outcomes of single and multicourses of IFN.
Patients and Methods
Among 573 consecutive HCV-infected patients in whom IFN monotherapy was induced between February 1987 and August 1992 at Toranomon Hospital, 313 were selected in the present study based on the following criteria: (1) patients naïve to IFN monotherapy; (2) patients infected with HCV genotype 1b alone; (3) patients with chronic hepatitis, without cirrhosis or HCC, as confirmed via biopsy examination within 6 months of enrollment; (4) patients not treated with IFN plus RBV combination therapy during follow-up time; (5) patients negative for hepatitis B surface antigen (radioimmunoassay, Dainabot, Tokyo, Japan), positive for anti-HCV (third-generation enzyme immunoassay, Chiron Corp., Emeryville, CA), and positive for HCV RNA qualitative analysis with PCR (nested polymerase chain reaction or Amplicor, Roche Diagnostic Systems, CA); (6) patients free of coinfection with human immunodeficiency virus; (7) patients not treated with antiviral or immunosuppressive agents within 6 months of enrollment; (8) lifetime cumulative alcohol intake <500 kg (mild to moderate alcohol intake); (9) patients free of other types of hepatitis, including hemochromatosis, Wilson's disease, primary biliary cirrhosis, alcoholic liver disease, and autoimmune liver disease; (10) patients without or with well-controlled diabetes; and (11) patients who consented to the study.
With regard to the clinical features of 313 patients at the start of the first course of IFN monotherapy, there were 223 men and 90 women aged 15-66 years with a median age of 47 years. The numbers of patients with fibrosis stages 1, 2, and 3 were 179, 107, and 27, respectively. The median ALT level was 138 IU/L (range, 24-636 IU/L), and the median platelet count was 17.4 × 104/μL (range, 8.9 × 104-39.2 × 104/μL). The median viremia level was 4.0 Meq/mL (range, <0.5-67.0 Meq/mL). The median follow-up time was 14.7 years (range, 0.1-20.1 years).
Furthermore, at the first course of IFN monotherapy, 222 patients (70.9%) received IFN-α alone; 83 patients (26.5%) received IFN-β alone; and the remaining 8 patients (2.6%) received a combination of IFN-α and IFN-β. A median IFN dose per day of 6 million units (MU) (range, 1-10 MU) was administered. As a whole, a median total dose of IFN of 525 MU (range, 22-3,696 MU) was administered during a median period of 23.9 weeks (range, 0.6 to 205.4 weeks). Patients mainly received IFN monotherapy, including initial aggressive induction therapy (every day within 8 weeks, followed by 3 times per week).
The study protocol was approved by the Human Ethics Review Committee of Toranomon Hospital.
The primary measure of efficacy of treatment was sustained virological response (SVR), defined as negative HCV RNA via qualitative analysis with PCR at 24 weeks after cessation of IFN therapy. Patients who achieved SVR after the first course of IFN monotherapy were classified as group A. Patients who did not achieve SVR after the first course of IFN monotherapy were classified into 2 groups based on whether they received other courses of IFN monotherapy. Patients who did not receive further courses of IFN monotherapy based on concerns about adverse effects, lack of time for treatment, physician recommendation based on the appearance of depression, and cardiopulmonary disease during and after the first course of IFN, or the lower levels of ALT, were classified as group B. Patients who received 2 or more courses of IFN monotherapy were classified as group C.
Blood samples were frozen at −80°C within 4 hours of collection and were not thawed until used for testing. HCV genotype was determined via PCR using a mixed primer set derived from nucleotide sequences of the NS5 region.14 In all cases, HCV-RNA viremia level was measured by branched DNA assay version 2.0 (Chiron Corp.) at commencement of therapy using frozen samples, and the results were expressed as 106genomic equivalents per milliliter (Meq/mL). The lower limit of the assay was 0.5 Meq/mL. Samples with undetectable levels using this quantitative assay (<0.5 Meq/mL) were also evaluated via HCV-RNA qualitative analysis with PCR (nested PCR or Amplicor) during and after therapy especially, and the results were expressed as positive or negative. The lower limit of the assay was 100 copies/mL.
Detection of Amino Acid Substitutions in Core Region.
We developed a simple and low-cost PCR method for detecting substitutions of aa 70 or aa 91 in HCV-CR of genotype 1b using mutation-specific primer as an alternative to the direct sequencing method. The major protein type was determined based on the relative intensity of the bands for wild (aa 70, arginine; aa 91, leucine) and mutant (aa 70, glutamine/histidine; aa 91, methionine) in agarose gel electrophoresis. If the intensities of the bands were similar, the case was regarded as competitive. The detection rate was 94.4%, the sensitivity was 10 KIU/mL using quantitative assay with PCR (Cobas Amplicor HCV monitor version 2.0 using the 10-fold dilution method), the reproducibility was high, and consistency with direct sequencing was 97.1% in positive cases.15 In this study, the pattern of arginine (wild) at aa 70 and leucine (wild) at aa 91 was evaluated as double wild-type, while the other patterns were nondouble wild-type. The mutation in this study refers to substitution from consensus sequence. In previous studies, HCV-J was considered as a prototype, and the aa substitution was evaluated by comparison with the consensus sequence prepared from 50 clinical trial samples.8, 16 In this study, the PCR genotyping could be performed in 232 patients; the remaining 81 patients could not be analyzed due to the lack of adequate serum samples obtained before treatment.
Liver Histopathological Examination.
Liver biopsy specimens were obtained percutaneously or at peritoneoscopy using a modified Vim Silverman needle with an internal diameter of 2 mm (Tohoku University style, Kakinuma Factory, Tokyo), fixed in 10% formalin, and stained with hematoxylin-eosin, Masson's trichrome, silver impregnation, and periodic acid-Schiff after diastase digestion. All specimens for examination contained 6 or more portal areas. Histopathological diagnosis was made by an experienced liver pathologist (H. K.) who was blinded to the clinical data. Chronic hepatitis was diagnosed based on histopathological assessment according to the scoring system of Desmet et al.17
Clinical and laboratory assessments were performed at least once every month before, during, and after treatment. Adverse effects were monitored clinically by careful interviews and medical examination at least once every month. Patient compliance with treatment was evaluated with a questionnaire. Blood samples were also obtained at least once every month before, during, and after treatment, and were also analyzed for ALT levels and HCV-RNA levels at various time points.
Follow-up time represented the time from the start of the first course of IFN treatment until death or until the last visit.
Diagnosis of HCC.
Patients were examined for HCC via abdominal ultrasonography every 3-6 months. If HCC was suspected based on ultrasonographic results, additional procedures such as CT, magnetic resonance imaging, abdominal angiography, and ultrasonography-guided tumor biopsy (if necessary), were used to confirm the diagnosis.
The χ2 test, Fisher exact probability test, and Mann-Whitney U test were used to compare background characteristics between groups. Multiple comparisons were examined by the Bonferroni test. Cumulative hepatocarcinogeneses were calculated using the Kaplan-Meier technique; differences between survival curves were tested using the log-rank test. Statistical analyses of hepatocarcinogenesis according to groups were calculated using the period from start of the first course of IFN monotherapy. Stepwise Cox regression analysis was used to determine independent predictive factors that were associated with hepatocarcinogenesis. We also calculated the OR and 95% CI. Potential predictive factors associated with hepatocarcinogenesis included the following 11 variables: age, sex, histological stage, viremia level, serum AST, serum ALT, platelet count, aa substitutions in HCV-CR, total IFN dose, total IFN duration, and group of treatment. Each variable was transformed into categorical data consisting of 2 simple ordinal numbers for univariate and multivariate analyses. Variables that achieved statistical significance (P < 0.05) or marginal significance (P < 0.10) on univariate analysis were tested using the multivariate Cox proportional hazard model to identify significant independent factors. Statistical comparisons were performed using the SPSS software (SPSS Inc., Chicago, IL). All P values of less than 0.05 by the 2-tailed test were considered significant.
Efficacy of IFN Monotherapy.
65 patients (20.8%) achieved SVR after the first course of IFN monotherapy (group A). Of 248 (79.2%) non-SVR patients after the first course of IFN, 112 (35.8%) did not receive a second course of IFN monotherapy (group B), while the remaining 136 (43.5%) received 2 or more courses of IFN monotherapy (group C). Of 136 patients in group C, 80 patients received 2 courses of IFN (21 of whom achieved SVR), 44 patients received 3 courses (6 of whom achieved SVR), 11 patients received 4 courses (2 of whom achieved SVR), and 1 patient received 6 courses (and did not achieve SVR). Thus, 29 patients in group C achieved SVR after multiple courses of IFN monotherapy.
In groups A and B, the median total duration of IFN was 24.1 weeks (range, 4.0-205.4 weeks) and 23.7 weeks (range, 2.9-75.1 weeks). The median total dose of IFN was 528 MU (range, 43-3,696 MU) and 498 MU (range, 72-870 MU). In the first, second, third, fourth, fifth, and sixth courses of IFN monotherapy in group C, the median total durations of IFN were 23.9 weeks (range, 0.6-136.4 weeks), 24.0 weeks (range, 1.3-313.7 weeks), 25.3 weeks (range, 3.1-198.1 weeks), 40.4 weeks (range, 21.0-86.3 weeks), 23.6 weeks, and 67.9 weeks, respectively. In the first, second, third, fourth, fifth, and sixth courses of IFN monotherapy in group C, the median total doses of IFN were 525 MU (range, 22-2,312 MU), 558 MU (range, 57-4005 MU), 522 MU (range, 28-3,477 MU), 565 MU (range, 363-1,080 MU), 708 MU, and 1,200 MU, respectively. The median cumulative total durations and cumulative total doses, which represented the cumulative total duration and total dose of every course of every patient of group C, were 65.6 weeks (range, 8.4-474.4 weeks) and 1,388 MU (range, 354-4,805 MU), respectively. The median periods free of IFN in group C were 3.6 years (range, 0.1-7.3 years). In conclusion, the median dose of IFN per week in group A, B, and C were 21.8 MU/week (range, 6.7-42.0 MU/week), 22.0 MU/week (range, 4.5-42.0 MU/week), and 21.9 MU/week (range, 3.7-43.9 MU/week), respectively.
Clinical Features of Patients and Cumulative Hepatocarcinogenesis Rates According to Study Groups.
The clinical features of patients in groups A, B, and C, at the start of the first IFN monotherapy are summarized in Table 1. The age of patients of group B was significantly higher than those of group A (P = 0.009; Bonferroni test) and group C (P = 0.007; Bonferroni test). Viremia levels in group A were significantly lower than those in group B (P < 0.001; Bonferroni test) and group C (P < 0.001; Bonferroni test). Fibrosis stage of group A was significantly milder than those of group B (P = 0.006; Bonferroni test) and group C (P = 0.009; Bonferroni test). There were no other significant differences in clinical features at the start of IFN therapy among the 3 groups.
|Group A (n = 65)||Group B (n = 112)||Group C (n = 136)|
|Age (years)*||44 (15–64)†||51 (23–66)||45 (22–63)‡|
|Viremia level (Meq/mL)*||0.6 (<0.5–45.0)||5.9 (<0.5–67.0)§||5.3 (<0.5–57.0)∥|
|Fibrosis stage (F1/F2/F3)||49/14/2||54/50/8¶||76/43/17#|
|AST (IU/L)*||83 (16–198)||74 (22–398)||75 (24–400)|
|ALT (IU/L)*||153 (24–416)||120 (38–636)||138 (50–594)|
|Platelet count (×104/μL)*||18.7 (9.7–31.0)||17.1 (9.7–39.2)||17.0 (8.9–31.2)|
|Core region (double wild/nondouble wild/ND)*||10/15/5||31/44/7||41/71/8|
During follow-up, 1 (1.5%), 17 (15.2%), and 15 (11.0%) patients developed HCC in groups A, B, and C, respectively. In groups A, B, and C, the cumulative hepatocarcinogenesis rates were 2.3%, 11.5%, and 0.8%, respectively, at the end of 5 years; 2.3%, 25.3%, and 7.2%, respectively, at the end of 10 years; and 2.3%, 33.0%, and 25.6%, respectively, at the end of 15 years. The rates were significantly different among the 3 groups (P < 0.001; Log-rank test) (Figure 1). In particular, the rates in group B were significantly higher than in group C (P < 0.001; Log-rank test) and group A (P < 0.001; Log-rank test), and the rates in group C were significantly higher than group A (P = 0.037; Log-rank test).
Hepatocarcinogenesis Rates According to aa Substitutions of HCV-CR.
During follow-up, 5 of 82 patients (6.1%) and 18 of 130 patients (13.8%) developed HCC in double wild-type and nondouble wild-type, respectively. In double wild-type and nondouble wild-type, the cumulative hepatocarcinogenesis rates were, respectively, 1.6% and 2.6% at the end of 5 years; 3.4% and 12.3% at the end of 10 years; and 11.3% and 23.5% at the end of 15 years. The rates in double wild-type of HCV-CR were significantly lower than those in nondouble wild-type (P = 0.036; log-rank test) (Fig. 2).
Predictive Factors Associated with Hepatocarcinogenesis via Multivariate Analysis.
We then analyzed the data for the whole population sample to determine those factors that could predict hepatocarcinogenesis. Univariate analysis identified 6 parameters that tended to or significantly correlated with carcinogenesis: age (P < 0.001), fibrosis stage (P < 0.001), platelet count (P < 0.001), group (P < 0.001), viremia level (P = 0.018), and aa substitution in HCV-CR (P = 0.036). These factors were entered into multivariate analysis, which identified 3 parameters that tended to or significantly influenced carcinogenesis independently: fibrosis stage (P < 0.001), aa substitutions in HCV-CR (P = 0.008), and group (P = 0.056) (Table 2).
|Factors||Category||Odds Ratio (95% CI)||P Value|
|Fibrosis stage||1: F1, F2||1|
|2: F3||10.2 (3.65–28.5)||<0.001|
|Amino acid substitutions||1: double-wild||1|
|in the core region||2: nondouble-wild||5.92 (1.58–22.2)||0.008|
|Group||1: A, C||1|
|2: B||2.75 (0.98–7.76)||0.056|
We also analyzed the data for 219 patients, except for 94 patients who achieved SVR, to determine those factors that could predict hepatocarcinogenesis. Univariate analysis identified 5 parameters that tended to or significantly correlated with carcinogenesis: fibrosis stage (P < 0.001), platelet count (P < 0.001), age (P = 0.001), group (P = 0.008), and aa substitution in HCV-CR (P = 0.028). These factors were entered into multivariate analysis, which identified 2 parameters that significantly influenced carcinogenesis independently: fibrosis stage (P < 0.001) and aa substitution in HCV-CR (P = 0.017) (Table 3).
|Factors||Category||Odds Ratio (95% CI)||P Value|
|Fibrosis stage||1: F1, F2||1|
|2: F3||6.50 (2.39–17.6)||<0.001|
|Amino acid substitutions||1: double-wild type||1|
|in the core region||2: nondouble wild-type||4.65 (1.32–16.4)||0.017|
Hepatocarcinogenesis Rates in Group C According to HCV-CR and ALT Levels.
In group C, the hepatocarcinogenesis rates were evaluated according to the ALT levels at the start of IFN. For this purpose, we selected 112 patients (82.4%) from group C in whom HCV-CR could be evaluated. In double wild-type, the hepatocarcinogenesis rates in patients with ALT levels below 1.5 (<75 IU/L) and above 1.5 (>75 IU/L) times the upper limit of normal (6-50 IU/L) were 0% (0/6 patients) and 8.6% (3/35 patients), respectively. In nondouble wild-type, the hepatocarcinogenesis rates in patients with ALT levels below 1.5 and above 1.5 times the upper limit of normal was 0% (0/7 patients), and 15.6% (10/64 patients), respectively (Table 4). In conclusion, regardless of whether aa substitutions in HCV-CR are present or not, lower hepatocarcinogenesis rates were noted in patients with ALT levels below 1.5 the upper limit of normal (0%) than in other patients (13.1%), but they did not achieve statistical significance on univariate analysis.
|ALT Level (IU/L)*|
|Nondouble wild-type||0% (0/7)||14.3% (2/14)||13.3% (4/30)||20.0% (4/20)|
|Double wild-type||0% (0/6)||16.7% (1/6)||5.6% (1/18)||9.1% (1/11)|
In group C, the hepatocarcinogenesis rates were also evaluated according to the mean ALT levels at the IFN-free period. For this purpose, we selected 76 consecutive patients (55.9%) from group C in whom ALT levels were closely monitored. In double wild-type, the hepatocarcinogenesis rates in patients with ALT levels below 4 (<200 IU/L) and above 4 (>200 IU/L) times the upper limit of normal were 0% (0/26 patients) and 50% (1/2 patients), respectively. In nondouble wild-type, the hepatocarcinogenesis rates in patients with ALT levels below 1.5 (<75 IU/L), from 1.5 to 2 (75-100 IU/L), from 2 to 4 (100-200 IU/L), and above 4 (>200 IU/L) times the upper limit of normal were 0% (0/13 patients), 33.3% (3/9 patients), 22.7% (5/22 patients), and 25.0% (1/4 patients), respectively (Table 5). In conclusion, regardless of whether aa substitutions in HCV-CR are present or not, significantly lower hepatocarcinogenesis rates were noted in patients with ALT levels below 1.5 times the upper limit of normal (0%) than in other patients (18.9%) (P = 0.027). In particular, significantly higher hepatocarcinogenesis rates were noted in patients of nondouble–wild-type with ALT levels above 1.5 times the upper limit of normal (25.7%) than in other patients (2.4%) (P = 0.004).
|ALT level (IU/L)*|
|Nondouble wild-type||0% (0/13)||33.3% (3/9)||22.7% (5/22)||25.0% (1/4)|
|Double wild-type||0% (0/10)||0% (0/4)||0% (0/12)||50.0% (1/2)|
Despite numerous lines of epidemiological evidence connecting HCV infection and the development of HCC, it remains controversial whether HCV itself plays a direct or indirect role in the pathogenesis of HCC.12 It is evident that the HCV-CR has oncogenic potential through the use of transgenic mice,13 but its clinical impact on hepatocarcinogenesis is still unclear. Our study identified that cumulative hepatocarcinogenesis rates of double wild-type HCV-CR, as a predictor of virological response for PEG-IFN plus RBV therapy, were significantly lower than those of nondouble wild-type. We speculate that the resistant cases for treatment might reasonably lead to HCC. To our knowledge, this is the first report to support the findings of oncogenic potential via HCV-CR from the clinical aspect. Previous reports identified PA28γ-dependent pathway as one of the mechanisms of HCV-associated hepatocarcinogenesis. Moriishi and colleagues 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.18, 19 Furthermore, HCV core protein also enhanced the binding of liver X receptor α/retinoid X receptor α to liver X receptor response element in the presence of PA28γ.19 Thus, it is reported that PA28γ plays a crucial role in the development of HCV-associated steatogenesis and hepatocarcinogenesis. Further studies should be performed to connect evidence from animal model studies and the clinical impact of aa substitution in HCV-CR on hepatocarcinogenesis.
Viral factors associated with hepatocarcinogenesis in patients infected with HCV are still incompletely investigated. Ogata et al. reported that HCV genotype 1b strains might be associated with HCC on the basis of the secondary structure of an amino-terminal portion of the HCV NS3 protein.20 Giménez-Barcons et al. reported that high aa variability within the NS5A of HCV might be associated with HCC in patients with HCV-1b–related cirrhosis.21 In the present study, we could not investigate the clinical impact of the other region on hepatocarcinogenesis, except for the HCV-CR. Further studies should be performed to investigate the clinical impact of the other region of HCV on hepatocarcinogenesis.
Patients who fail to achieve SVR after single-course IFN should receive multicourse IFN at the time of ALT relapse at certain intervals. Based on previous reports showing increased incidence of HCC in 5 years or more after IFN therapy in transient biochemical responders, it is important to normalize ALT levels via multicourse IFN monotherapy at certain intervals.11, 22 We reported previously that results of multicourse IFN showed a 0% hepatocarcinogenesis rate in patients with ALT levels below 75 IU/L at the IFN-free periods, emphasizing the importance of keeping low ALT levels at such periods with respect to suppression of hepatocarcinogenesis.11 Furthermore, hepatocarcinogenesis rates according to HCV-CR and ALT levels during the IFN-free period were also evaluated in this study. In double wild-type, the rates in patients with ALT levels below 200 IU/L and above 200 IU/L were 0% and 50%, respectively. In nondouble wild-type, the rates in patients with ALT levels below 75 IU/L and above 75 IU/L were 0% and 25.7%, respectively. Thus, significantly higher hepatocarcinogenesis rates were noted in patients of nondouble wild-type with ALT levels above 75 IU/L than in other patients. In particular, in multicourse IFN therapy in nondouble wild-type, we emphasize the importance of reducing the risk of hepatocarcinogenesis by the mean ALT during the IFN-free period below 1.5 times the upper limit of normal.
It is unclear whether ALT levels during the IFN-free period might be more important than those at the start of IFN. In the present study, at the start of IFN, lower hepatocarcinogenesis rates were noted in patients with ALT levels below 1.5 the upper limit of normal compared with other patients, but they did not achieve statistical significance on univariate analysis. During the IFN-free period, significantly lower hepatocarcinogenesis rates were noted in patients with mean ALT levels below 1.5 times the upper limit of normal compared with other patients. Thus, in multicourse IFN therapy, especially in nondouble wild-type, we emphasize the importance of reducing the risk of hepatocarcinogenesis via ALT levels below 1.5 times the upper limit of normal during the IFN-free period rather than at the start of IFN. Further studies should be conducted in the future to confirm this finding.
To our knowledge, our study is the first to report the hepatocarcinogenesis rates for a long-term follow-up period of 15 years in IFN monotherapy. Previous studies have shown that sex, age, fibrosis stage, and IFN regimen are important pretreatment predictors of hepatocarcinogenesis.11, 23–25 In the present study, a more progressive fibrosis stage as host factor, nondouble wild-type of HCV-CR as viral factor, and group B (non-SVR after single-course IFN) as treatment-related factor were associated with higher hepatocarcinogenesis rates in the whole population sample. Even if we also analyzed non-SVR patients, multivariate analyses similarly identified more progressive fibrosis stage and nondouble wild-type of HCV-CR that significantly influenced hepatocarcinogenesis independently. Hence, we assess that the risk of HCC is not necessarily secondary to the lack of response to IFN therapy rather than aa substitution. We conclude that hepatocarcinogenesis seems to be based on a dynamic tripartite interaction of virus, host, and treatment regimen. Further understanding of the complex interaction between these factors should facilitate the development of more effective therapeutic regimens. In Japan, only 5 years had elapsed since the induction of IFN-α2b plus RBV combination therapy (especially, only 2 years in PEG-IFN-α2b plus RBV) based on the Japanese Government Health Insurance system, so we could not exactly evaluate the long-term efficacy of combination therapy as a treatment-related factor of hepatocarcinogenesis in this study. Further studies that include patients treated not only with IFN monotherapy but also with RBV combination therapy should be performed in the future.
The relationship between the development of cirrhosis and HCC is still unclear. We investigated liver fibrosis stage of 13 patients who underwent partial hepatectomy for HCC in this study. Interestingly, 8 of 13 patients (61.5%) developed HCC in the absence of cirrhosis (5 patients of fibrosis stage 2, 3 patients of fibrosis stage 3). As a whole, it is regrettable that we could not exactly evaluate how frequently HCC occurs in the absence of cirrhosis. Further studies based on all patients, whether or not they develop HCC, should be performed to investigate the relationship between the development of cirrhosis and HCC.
In conclusion, aa substitutions in the HCV-CR are the primary predictor of hepatocarcinogenesis. In particular, in multicourse IFN therapy in nondouble wild-type as a pretreatment negative predictor of SVR for PEG-IFN plus RBV combination therapy, we emphasize the importance of reducing the risk of hepatocarcinogenesis via mean ALT levels below 1.5 times the upper limit of normal during the IFN-free period. Furthermore, IFN monotherapy should be recommended as a therapeutic regimen to reduce the risk of hepatocarcinogenesis in patients unsuitable for PEG-IFN plus RBV combination therapy. Large-scale prospective studies should be conducted in the future to confirm this finding.
- 10Predictive factors of early and sustained responses to peginterferon plus ribavirin combination therapy in Japanese patients infected with hepatitis C virus genotype 1b: amino acid substitutions in the core region and low-density lipoprotein cholesterol levels. J Hepatol 2007; 46: 403–410., , , , , , et al.
- 11Long-term follow-up of interferon monotherapy in 454 consecutive naive patients infected with hepatitis C virus: multi-course interferon therapy may reduce the risk of hepatocellular carcinoma and increase survival. Scand J Gastroenterol 2005; 40: 688–696., , , , , , et al.
- 20Identification of hepatitis C virus subtype 1b strains that are highly, or only weakly, associated with hepatocellular carcinoma on the basis of the secondary structure of an amino-terminal portion of the HCV NS3 protein. J Clin Microbiol 2003; 41: 2835–2841., , , , , , et al.
- 23Effect of interferon therapy on hepatocellular carcinogenesis in patients with chronic hepatitis type C: a long-term observation study of 1643 patients using statistical bias correction with proportional hazard analysis. HEPATOLOGY 1999; 29: 1124–1130., , , , , , et al.