Potential conflict of interest: Nothing to report.
The interferon sensitivity determining region (ISDR) of the hepatitis C virus (HCV) and T-helper type 1 and type 2 (Th1/Th2) ratio were analyzed along with other host and viral factors for their ability to predict the response of patients with chronic hepatitis C to pegylated interferon alpha-2b (Peg-IFN) and ribavirin (RBV) combination therapy. A total of 120 chronic hepatitis C patients with genotype 1 HCV and high baseline viral loads who were to undergo combination therapy scheduled for 48 weeks were enrolled. Sustained virologic response (SVR) was achieved in 54 (45%) of the 120 patients. The pretreatment factors significantly associated with SVR by logistic regression analysis were ISDR mutant [odds ratio (OR) = 86.0, P = 0.0008], Th1/Th2 ratio ≤ 15.5 (OR = 9.6, P = 0.0021), body weight 59 kg, and neutrophil count 2,300/μL. A logistic regression model to estimate SVR before combination therapy was constructed using these four factors. Patients fell into three groups when plotted according to estimated and actual SVR rates: actual SVR rate was 91% (32/35) in the high sensitivity group, 41% (15/37) in the intermediate sensitivity group, and 15% (7/48) in the low sensitivity group. Rapid or early virological responses were seen in 80% of patients with high sensitivity and who achieved SVR but were found in only 40% of patients with intermediate or low sensitivity. Null- and very late virological responses were quite rare in the high sensitivity group. In conclusion, a logistic regression model that includes the sequence of ISDR of the HCV, Th1/Th2 ratio, body weight, and neutrophil count can be useful for accurately predicting actual SVR rate before combination therapy. (HEPATOLOGY 2008;48:1753-1760.)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Chronic infection with hepatitis C virus (HCV) can lead to chronic hepatitis and eventually liver cirrhosis and hepatocellular carcinoma.1 Administration of antiviral agents such as interferon (IFN) can eradicate HCV in some patients with chronic hepatitis C, and the risk of complicating hepatocellular carcinoma has been reported to decrease remarkably once this is achieved.2–6
HCV genotype and viral load are two major factors used to predict the response of patients with chronic hepatitis C to IFN. Patients who have genotype 1 HCV and high viral loads are relatively resistant to IFN therapy.7 Peg-IFN and RBV combination therapy is currently the first line of therapy for these cases.8 However, although the sustained virologic response (SVR) rate has been improved with the advent of combination therapy, it remains approximately 50%. The velocity of decrease in viral load during combination therapy is also a good indicator for predicting SVR; high SVR rates are predicted in rapid and early virological responders, whereas low SVR rates are predicted in late and nonvirological responders.9–12
It is considered beneficial to predict the response of patients with genotype 1 HCV and high viral load to Peg-IFN and RBV combination therapy before starting treatment because therapy can be long, costly, and have many side effects. However, prediction is often difficult in these patients because they have already been selected as being poor responders to IFN therapy by the major prediction factors of HCV genotype and viral load.
Amino acid substitutions in the interferon sensitivity determining region (ISDR), located in HCV nonstructural region 5A, have also been reported as useful for predicting the response of patients with genotype 1 to IFN therapy.13 Several host factors, such as immunological responses, are suggested to be associated with viral response as well.14–22 Among them, we chose the T-helper type 1 and type 2 (Th1/Th2) ratio as a representative host factor for analysis.
In the current study, we studied whether a combination of viral and host factors, including the presence of ISDR mutants and Th1/Th2 ratio, could predict response to Peg-IFN and RBV therapy in chronic hepatitis C patients with genotype 1 HCV and high viral load.
BMI, body mass index; EVR, early virologic response; HCV, hepatitis C virus; IFN, interferon; ISDR, interferon sensitivity determining region; Null-R, null-response; Peg-IFN, peginterferon-alpha-2b; RBV, ribavirin; RVR, rapid virological response; SVR, sustained virological response; Th1/Th2 ratio, T-helper type 1 and type 2 ratio; VLVR, very late virological response.
Patients and Methods
A total of 120 patients with chronic hepatitis C were treated with Peg-IFN and RBV combination therapy at Shinshu University Hospital and the 21 member hospitals of the Shinshu Interferon Treatment Research Group. The cohort included 65 men and 55 women, ranging from 17 to 75 years of age, who were registered prospectively from December 2004 to December 2005. All patients had HCV genotype 1b and had shown high viral load for at least 6 months. High viral load was defined as serum HCV RNA equal to or greater than 105 international units/mL as measured by quantitative Cobas Amplicor assays (Roche Diagnostics Co. Ltd, Tokyo, Japan). Of the 120 patients, 86 had undergone a liver biopsy before combination therapy, and seven of the 86 patients were diagnosed as having cirrhosis. Exclusion criteria for patients not eligible for Peg-IFN and RBV combination therapy were as follows: (1) pregnant women or women of child-bearing potential, nursing mothers, or male patients whose partner might become pregnant; (2) patients with anemia (hemoglobin concentration of 10 g/dL or less), leukopenia (1,500/μL or less), or thrombocytopenia (80,000/μL or less); (3) patients with depression; (4) patients with serious complications in the heart, kidneys, or lungs; (5) patients with autoimmune diseases, such as autoimmune hepatitis; (6) patients infected with hepatitis B virus or human immunodeficiency virus; and (7) patients with hypersensitivity to Peg-IFN or RBV.
This study was approved by the ethics committee of Shinshu University and performed in accordance with the internationally accepted ethical standards for human experimentation. The purpose and the protocol of this study were explained to all patients, and written informed consent was obtained from each participant.
Peg-IFN and RBV Combination Therapy.
Peg-IFN-α-2b (Schering-Plough K.K., Tokyo, Japan) was given in weekly doses adjusted to body weight according to manufacturer's instructions (45 kg or less; 60 μg/dose, 46 to 60 kg; 80 μg/dose, 61 to 75 kg; 100 μg/dose, 76 to 90 kg; 120 μg/dose, 91 kg or more; 150 μg/dose). Similarly, RBV (Schering-Plough K.K.) was given in daily doses adjusted to body weight according to manufacturer's instructions (60 kg or less; 600 mg/day, 61 kg to 80 kg; 800 mg/day, 81 kg or more; 1,000 mg/day). The duration of the combination therapy was set at a standard 48 weeks, but treatment extension was permitted to up to 72 weeks if the patient requested.
A rapid virologic response (RVR) was defined as undetectable serum HCV RNA at 4 weeks as measured by qualitative Cobas Amplicor assays (Roche Diagnostics Co. Ltd., Tokyo, Japan). An early virological response (EVR) was defined as detectable serum HCV RNA at 4 weeks but undetectable by 12 weeks, a late virological response was defined as serum HCV RNA detectable at 12 weeks but undetectable by 24 weeks, a very late virological response (VLVR) was defined as serum HCV RNA detectable at 24 weeks but undetectable by 48 weeks, and a null-response (Null-R) was defined as serum HCV RNA not becoming undetectable during the treatment course. An end of treatment response was defined as negative serum HCV RNA by the end of treatment. An SVR was defined as serum HCV RNA becoming undetectable during therapy and remaining so for at least 24 weeks afterwards. Responses other than SVR were regarded as non-SVR.
Achieved rates of Peg-IFN and RBV administration were calculated as the percentage of actual total dose administered of a standard total dose of 48 weeks calculated according to body weight before therapy.
Serological Tests for HCV, Hepatitis B Virus, and Human Immunodeficiency Virus.
Antibodies to HCV, hepatitis B virus surface antigen, and human immunodeficiency virus were measured using commercially available enzyme-linked immunosorbent assays (International Reagents Co., Kobe, Japan). Serum HCV RNA was determined using qualitative and quantitative Cobas Amplicor assays (Roche Diagnostics Co. Ltd., Tokyo, Japan), which amplify HCV RNA using reverse transcription polymerase chain reaction. HCV genotypes were determined according to the method reported by Ohno et al.23 Serum alanine aminotransferase and other relevant biochemical tests were performed using standard methods.
Serum Level of Ribavirin.
Serum level of ribavirin was measured using a validated liquid chromatography/ tandem mass spectrometric assay with a detection limit of 50 ng/mL.24 Serum samples were taken at 1 and 8 weeks after starting combination therapy.
Amino Acid Substitutions in the ISDR.
ISDR type was determined by the method reported by Enomoto et al.,13 in which the HCV-J strain of genotype 1b, as reported by Kato et al.,25 was used as the wild type. Briefly, the nucleotide sequence of ISDRs in the nonstructural 5A region was determined by direct sequencing of polymerase chain reaction amplified materials to deduce amino acid sequence. Wild-type ISDR was defined as having no amino acid substitutions, intermediate-type ISDR was defined as having one amino acid substitution, and mutant-type ISDR was defined as containing two or more amino acid substitutions.
The Th1/Th2 ratio in peripheral blood was determined using flow cytometry according to the method reported by Kawakami et al.26 Briefly, CD4-positive cells were extracted, then Th1 (IFN-γ+/interleukin-4−) and Th2 (IFN-γ−/interleukin-4+) cells were classified using monoclonal antibodies to IFN-γ and interleukin-4. The Th1/Th2 ratio was calculated as number of Th1 cells per number of Th2 cells.
The Mann-Whitney U-test was used to analyze continuous variables. Chi-squared and Fisher's exact tests were used for analysis of categorical data. Multivariate analysis was performed using a logistic regression model with stepwise method. Each cutoff point for continuous variables was decided by receiver operating characteristic curve analysis. A P-value of less than 0.05 was considered significant. Statistical analyses were performed using SPSS for Windows v16.0J (SPSS Inc, Chicago, IL).
Response Rate and Clinical Characteristics.
SVR was achieved in 54 (45%) of the 120 patients enrolled in the current study. In total, 18 (15%) patients elected to extend treatment to up to 72 weeks, although SVR rate was similar between patients with (44%) and without (45%) the extension. Discontinuation of Peg-IFN and RBV combination therapy during treatment course was recorded in four (7%) of the 54 patients with SVR and 21 (32%) of the 66 patients with non-SVR. Of the 21 non-SVR patients who discontinued combination therapy, nine were because of side effects and 12 because of insufficient effects, namely serum HCV RNA remaining detectable at 24 weeks.
Factors Associated with SVR.
Pretreatment factors that could be associated with responses to Peg-IFN and RBV combination therapy were compared between patients with and without SVR in Table 1. Patients with SVR tended to be younger than those with non-SVR and who were male. Body weight was higher in SVR patients, but body mass index (BMI) did not differ between the two groups. Median counts of white blood cells, neutrophils, and platelets and median concentrations of creatinine and hemoglobin were significantly higher in patients with SVR than in those without. Mutant ISDR was more prevalent in patients with SVR, but the Th1/Th2 ratio was significantly lower.
Table 1. Comparison of Pretreatment Factors Between Patients With and Without SVR
SVR (n = 54)
Non-SVR (n = 66)
Data are expressed as median (range).
W, wild; I, intermediate; M, mutant; UD, undermined.
Predictive factors measured during treatment were also compared between patients with and without SVR (Table 2). Serum concentrations of RBV at 1 and 8 weeks of therapy did not differ between the two groups. Total administered dose of Peg-IFN and RBV was significantly higher in patients with SVR. RVR and EVR were more prevalent in patients with SVR, whereas VLVR and Null-R were more prevalent in patients without.
Table 2. Comparison of Treatment Factors and Virological Responses During Peg-IFN and RBV Therapy Between Patients With and Without SVR
SVR (n = 55)
Non-SVR (n = 66)
Data are expressed as median (range).
Achieved administration rate for Peg-IFN and RBV was calculated as the percentage of actual dose administered of the scheduled dose for 48 weeks.
RVR, rapid virological response; EVR, early virological response; LVR, late virological response; VLVR, very late virological response; Null-R, null response.
Factors that were significantly associated with SVR by univariate analysis were then analyzed by multivariate analysis. Both pretreatment and treatment factors were analyzed together to select the pretreatment prediction factors that were independent from the treatment prediction factors. Cutoff points for continuous data were determined by receiver operating characteristic analysis and were as follows: 57 years old, body weight 59 kg, BMI 23 kg/m2, creatinine 0.75 mg/dL, white blood cells 4,200 cells/μL, neutrophils 2,300 cells/μL, hemoglobin 15.0 g/dL, platelets 135,000 cells/μL, HCV RNA 6.0 × 105 international units/mL, Th1/Th2 ratio 15.5, total Peg-IFN dose 2900 μg, total RBV dose 182 g, achieved rate of Peg-IFN 73% of target amount, and achieved rate of RBV 79% of target amount. The seven factors shown in Table 3 were then evaluated by logistic regression analysis with stepwise method, indicating that mutant ISDR, Th1/Th2 ratio 15.5 or lesser, body weight 59 kg or greater, and neutrophils 2,300 cells/μL or greater were significantly associated with SVR among pretreatment factors. The odds ratio of mutant ISDR was as high as 86.0, and the odds ratios of the remaining three pretreatment factors all fell between 5.0 and 10.0. The positive predictive values of ISDR mutant, Th1/Th2 ratio 15.5 or lesser, body weight 59 kg or greater, and neutrophil count 2300 cells/μL or greater were 82.8%, 61.8%, 63.1%, and 54.1%, respectively, and negative predictive values were 67.0%, 69.2%, 76.4%, and 64.4%, respectively. As for treatment factors, RVR, EVR, and a higher dose of Peg-IFN were also found to be factors predicting SVR.
Table 3. Multivariate Logistic Regression Analysis for Factors Associated with SVR
Cutoff value for each factor was determined by receiver operating characteristic curve (ROC) analysis.
Th1/Th2 ratio ≤ 15.5
Body weight ≥ 59kg
Neutrophil count ≥ 2,300/μL
Virological response during therapy
Rapid virological response
Early virological response
Total dose of Peg-IFN ≥ 2,900 μg
Pretreatment Prediction of SVR by Logistic Model.
A logistic regression model for predicting SVR was constructed using the four pretreatment factors significantly associated with SVR by multivariate analysis:
Figure 1 shows the distribution of patients according to predicted SVR rates, which are well correlated with actual SVR rates. The patients were further divided into three groups: a high-sensitivity group for patients whose predicted SVR rates were in the upper one third (>66%), an intermediate group for SVR rates in the middle one third (33%–66%), and a low-sensitivity group for rates in the lower one third (<33%). The actual SVR rates were 91% (32/35), 41% (15/37), and 15% (7/48) in the high, intermediate, and low sensitivity groups, respectively.
The actual SVR rates were then compared among the three groups classified according to standard doses of Peg-IFN and RBV administration (Fig. 2). The first group consisted of patients who took lesser amounts of treatment (Peg-IFN dose less than 73% and RBV dose less than 79% of target amounts). The second group received Peg-IFN over 73% or RBV over 79% of target amounts. The third group consisted of patients who received both Peg-IFN over 73% and RBV over 79% of target amounts. SVR rates were similarly low among patients with low sensitivity and high among patients with high sensitivity, but rose with increases of Peg-IFN and RBV doses received in patients with intermediate sensitivity.
The SVR rates in patients with RVR, EVR, late virological response, VLVR, and Null-R were 85% (22/26), 68% (23/34), 39% (9/23), 0% (0/5), and 0% (0/32), respectively. Distributions of actual virological responses during therapy are shown according to the estimated sensitivity of low, intermediate, and high in Fig. 3, and are significantly associated with the estimated sensitivities. RVR and EVR were seen in 80% of patients with high sensitivity, but were less than 40% in patients with intermediate or low sensitivity. Null-R and VLVR were quite rare in the high-sensitivity group. All patients with RVR and EVR in the high-sensitivity group achieved SVR.
Besides HCV genotype and viral load, several factors, including age, sex, race, BMI, HCV mutations, and host immunological parameters, have been reported to be associated with SVR rates in patients with chronic hepatitis C treated with Peg-IFN and RBV combination therapy. However, virological responses during therapy, such as RVR and EVR, are now widely used for predicting final virological response because such treatment factors have even higher predictive value.9–12 Nonetheless, it is obvious that predictions made before administration of therapy are more desirable than those done during treatment course if accuracy of prediction is comparable. We therefore planned the current study to find such a way to predict SVR before starting Peg-IFN and RBV combination therapy in patients already known to be resistant.
Several regions in the HCV genome have been reported to be associated with sensitivity to interferon therapy. Enomoto et al.13 reported that a higher number of amino acid substitutions in the ISDR (NS5A, a.a. 2209–2248) were strongly associated with a favorable response to IFN-α monotherapy in patients with genotype 1 HCV.13 It is postulated that the NS5A protein, in which the ISDR exists, has transcriptional activation functions and represses interferon-induced gene expression.27 The ISDR overlaps a putative acidic amino acid region that confers transcriptional activity.28 Akuta et al.29 reported that amino acid substitutions of R by Q at a.a. 70 or L by M at a.a. 91 in the core region were significantly frequent in patients who showed a null or weak response to combination treatment of 48 weeks. As such, we chose the ISDR among all HCV genetic factors for analysis because it has already been well characterized.30
Immunological backgrounds are known to be associated with response to IFN therapy because cellular immune functions are essential to eliminate HCV-infected hepatocytes. Masaki et al.15 reported that a lower Th1/Th2 ratio before IFN monotherapy was a significant host factor for predicting long-term virological response in Japanese patients with chronic hepatitis C. Lee et al.31 reported that high baseline sCD30 levels predicted an early and sustained virological response to IFN and RBV therapy, and suggested that therapy might be more effective in patients with a predominant T2 profile. Lagging et al.32 reported that low levels of a 10-kDa IFN-γ inducible protein predicted rapid and sustained virological response in patients with genotype 1 HCV treated with Peg-IFN and RBV combination therapy. Taken together, these results indicate that an imbalance of Th1 and Th2 subsets before IFN therapy is possibly associated with long-term therapy outcome. In the current study, we chose the Th1/Th2 ratio in peripheral blood as an immunological marker for predicting SVR because identification of helper T cell subpopulations at the cellular level has become practical with the development of intracellular cytokine assays using flow cytometry.
Of the virological and host factors analyzed in the current study, mutant ISDR, Th1/Th2 ratio 15.5 or less, body weight 59 kg, and neutrophil count 2300 cells/μL were selected as significant pretreatment factors predicting a higher rate of SVR. Although our study suggests that higher body weight is a favorable predictor for SVR in the Japanese, several studies on Caucasians have shown that higher body weight or BMI results in a lower SVR rate.33, 34 This difference may be attributed to a difference in average body weight among the studies; whereas median body weight was over 70 kg and patients with body weights of less than 59 kg were quite rare in studies reported from the United States and Europe, the median body weight in our study was 60.7 kg, and patients with body weights of less than 59 kg accounted for 45.8% of our cohort. According to manufacturer's instructions, patients with body weights equal to or less than 60 kg received 80 μg of Peg-IFN-α-2b and 600 mg RBV as initial doses in the current study. It is possible that such doses were insufficient to achieve a high SVR rate despite having been adjusted accordingly. This possibility is further supported by our result that the distribution of BMI did not differ between patients with and without SVR.
A higher neutrophil count was also a significant pretreatment factor predicting a greater likelihood of SVR. Patients with chronic hepatitis C who have advanced fibrosis tend to show lower counts of neutrophils and platelets, which can interfere with administration of Peg-IFN, another significant treatment factor identified in this study. Indeed, pretreatment counts of neutrophils were significantly higher in patients who received sufficient (73% or more) Peg-IFN doses than in those who did not.
Interestingly, a lower Th1/Th2 ratio predicted a higher SVR rate in our study, contrary to the common knowledge that a stronger Th1 response is important to eradicate HCV; Shinohara et al.35 reported that a higher increase in Th1 response during the early phase of interferon therapy was associated with higher SVR rates. However, these patients also showed a lower Th1/Th2 ratio before starting therapy than those without SVR. Thus, our result that a lower Th1/Th2 ratio predicts a favorable response does not necessarily refute the importance of the Th1 response in eradicating HCV. Further studies are required to clarify the significance of the Th1/Th2 ratio.
The logistic regression model for predicting SVR in this study yielded three patient groups classified according to predicted SVR rate. The actual SVR rates were 91% in the high sensitivity group, 41% in the intermediate sensitivity group, and 15% in the low sensitivity group, and all were well correlated with predicted SVR rates. Earlier clearance of HCV viremia during combination therapy has been reported to be a good indicator for higher SVR rate. As expected, rapid and early virological responses were seen in 80% of patients with high sensitivity but were found in only 40% of patients with intermediate or low sensitivity. Null or very late virological responders were quite rare in the high-sensitivity group. Because all rapid and early virological responders achieved SVR, patients who are judged to have high sensitivity are strongly recommended to take Peg-IFN and RBV combination therapy.
It is noteworthy that actual SVR rates were similarly low in the low-sensitivity group and high in the high-sensitivity group irrespective of total doses of Peg-IFN and RBV administered, but were significantly associated with total dose in the intermediate-sensitivity group. Thus, efforts to maintain at least standard administration doses of Peg-IFN and RBV are important to achieve higher SVR rates in patients who have intermediate sensitivity to combination therapy.
Several pretreatment factors other than ours have been reported to be associated with virological response to IFN and RBV combination therapy in patients with genotype 1b HCV infection, including serum levels of low-density lipoprotein cholesterol and insulin resistance. However, the results obtained here are valuable because the estimated sensitivities to combination therapy are very closely associated with actual sensitivities. Our algorithm cannot be applied to other populations directly because clinical backgrounds, such as distribution of body weight, differ globally. However, a number of mutations in the ISDR sequence of the hepatitis C virus have been shown to be associated with response to IFN therapy in a worldwide meta-analysis reported by Pascu et al.30 Immunological factors, such as Th1/Th2 imbalance, have also been reported to be associated with treatment response in the world.16, 17 As such, the current study is applicable on a global scale because it clearly shows that a combination of viral and host factors, including those of immunological nature, are effective for predicting the response to combination therapy before it is started.
In conclusion, in Japanese patients from Nagano, SVR rates of Peg-IFN and RBV combination therapy can be accurately predicted using the pretreatment factors of ISDR mutations, Th1/Th2 ratios, body weights, and neutrophil counts. This being established, future prospective trials are required to validate our results in other regions of Japan and in other countries.
We thank Toyo Amaki and Asami Yamazaki for their technical assistance, and Trevor Ralph for his editorial assistance.
The Nagano Interferon Treatment Research Group is consisted of following members; Drs. Kaoru Arai, Noriko Misawa, Koji Orii, (Internal Medicine, Shinshu University School of Medicine, Matsumoto), Dr. Yukio Gibo (Gibo Liver Clinic, Matsumoto), Dr. Yasuharu Imai (Gastroenterology, Tokyo Medical University Hospital, Tokyo), Dr. Hidetomo Muto (Nagano Municipal Hospital, Nagano), Drs. Shuichi Wada, Hiromitsu Mori, (Nagano Red Cross Hospital, Nagano), Dr. Kazuhiro Funase (Kamijo Kinenn Hospital, Matsumoto), Dr. Chiharu Miyabayashi (Chikuma Central Hospital, Chikuma), Dr. Yuriko Koike (Kawanakajima Clinic, Nagano), Dr. Akihiro Iijima (Kiso Prefecture Hospital, Kisofukushima), Dr. Masato Takamatsu (Saku Central Hospital, Usuda), Drs. Toshihisa Tsukadaira and Akinori Kitahara (Kenwakai Hospital, Iida), Drs. Takeshi Sodeyama and Masakazu Kobayashi (National Chushin Matsumoto Hospital, Matsumoto), Dr. Kiyoshi Furuta, (Matsumoto Hospital, Matsumoto), Dr. Haruhiko Imai (Yodakubo Hospital, Nagato-machi), Dr. Hidetoshi Yoda (NTT Nagano Hospital, Nagano), Dr. Seiichi Usuda (Aizawa Hospital, Matsumoto), Dr. Yoshiyuki Nakatsuji (Nagano Hospital, Ueda), Dr. Akihiko Urushihara (Tatsuno Hospital, Tatsuno-machi), Dr. Wataru Okiyama (Suwa Red Cross Hospital, Suwa), Dr. Atsushi Maruyama (Ina Central Hospital, Komagane), Dr. Takahiro Yamaura (Iida Municipal Hospital, Iida).