Journal of Gastroenterology

, Volume 49, Issue 11, pp 1485–1494

Baseline factors and very early viral response (week 1) for predicting sustained virological response in telaprevir-based triple combination therapy for Japanese genotype 1b chronic hepatitis C patients: a multicenter study

Authors

    • Division of Gastroenterology and HepatologyShinmatsudo Central General Hospital
  • Hidenori Toyoda
    • Department of GastroenterologyOgaki Municipal Hospital
  • Akihito Tsubota
    • Institute of Clinical Medicine and ResearchJikei University School of Medicine
  • Tatsuya Ide
    • Division of Gastroenterology, Department of MedicineKurume University School of Medicine
  • Koichi Takaguchi
    • Department of HepatologyKagawa Prefectural Central Hospital
  • Keizo Kato
    • Division of Gastroenterology and HepatologyShinmatsudo Central General Hospital
  • Masaki Kondoh
    • Department of Life Cycle ManagementRoche Diagnostics K.K
  • Kazuhiro Matsuyama
    • Department of Life Cycle ManagementRoche Diagnostics K.K
  • Takashi Kumada
    • Department of GastroenterologyOgaki Municipal Hospital
  • Michio Sata
    • Division of Gastroenterology, Department of MedicineKurume University School of Medicine
Original Article—Liver, Pancreas, and Biliary Tract

DOI: 10.1007/s00535-013-0918-7

Cite this article as:
Shimada, N., Toyoda, H., Tsubota, A. et al. J Gastroenterol (2014) 49: 1485. doi:10.1007/s00535-013-0918-7

Abstract

Background

Genetic polymorphisms near Interleukin 28B (IL28B) (rs8099917) and a rapid virological response (RVR) have been reported as predictors for a sustained virological response (SVR) to telaprevir (TVR)-based triple combination therapy. However, the association between SVR and viral kinetics earlier than week 4 after initiation of therapy remains unclear. Thus, we evaluated the SVR prediction ability of baseline factors and reduced hepatitis C virus (HCV) RNA levels at week 1 after the initiation of TVR-based therapy in Japanese genotype-1b chronic hepatitis C (CHC) patients.

Methods

A total of 156 Japanese CHC patients received a 24-week regimen of TVR-based therapy. Baseline factors and reduction in HCV RNA levels at weeks 1 and 4 after the initiation of therapy were analyzed for SVR prediction.

Results

Multiple logistic regression analysis for SVR in TVR-based therapy identified the IL28B TT genotype, a reduction of ≥4.7 log10IU/mL in HCV RNA levels at week 1, RVR, and treatment-naïve/relapse. Whereas the SVR rate was higher than 90 % regardless of the reduction in HCV RNA levels at week 1 in patients with the TT genotype, a reduction of ≥4.7 log10IU/mL in HCV RNA levels at week 1 was the strongest predictor of SVR in patients with the non-TT genotype, as determined by multiple logistic regression analysis (P = 0.0043).

Conclusions

The IL28B TT genotype is the most important baseline factor for predicting SVR, and a ≥4.7 log10IU/mL reduction in HCV RNA at week 1 is a useful very early on-treatment predictor of SVR, especially in the non-TT genotype.

Keywords

Chronic hepatitis CReduction in HCV RNA at week 1TelaprevirIL28B

Introduction

In 2011, the first-generation direct-acting antiviral agents telaprevir (TVR) and boceprevir (BOC) were approved for treatment of chronic hepatitis C (CHC) patients with hepatitis C virus (HCV) genotype 1 in the United States (US), Canada, and the European Union (EU). Triple combination therapy with TVR or BOC, PEG-interferon (PEG-IFN), and ribavirin (RBV) is the current standard of care for genotype 1 CHC patients [1]. In Japan, TVR, which is a nonstructural (NS) 3/4A serine protease inhibitor, was approved in September 2011 and has been marketed since November 2011. In treatment-naïve genotype 1 CHC patients, TVR-based triple combination therapy for a shortened period was reported to remarkably improve the rate of sustained virological response (SVR) compared with PEG-IFN and RBV alone [24]. In treatment-experienced patients, the effect of TVR-based triple combination therapy reportedly depends on the response to PEG-IFN and RBV combination therapy [516].

Pivotal genome-wide association studies have found that genetic variations near the interleukin 28B (IL28B) gene (rs8099917 and rs12979860) are strongly associated with the treatment outcome of PEG-IFN and RBV combination therapy [1719]. We previously confirmed that the IL28B single-nucleotide polymorphism (SNP) genotype was the strongest factor contributing to SVR in PEG-IFN and RBV combination therapy [2023]. These genetic variations appear to be strong predictors of SVR to a 24-week regimen of TVR-based triple therapy, as well as PEG-IFN and ribavirin combination therapy [7, 11, 1416, 24].

Two guidelines for treatment of genotype 1 CHC patients, which were based on the results of the clinical trials of a 24-week regimen of TVR-based triple therapy for Japanese patients [4, 10], provided recommendations for patient selection for TVR-based therapy [25]. Study Group for the Standardization of Treatment of Viral Hepatitis Including Cirrhosis published by the Ministry of Health, Labour and Welfare of Japan, and Japan Society of Hepatology guidelines provided recommendations of a 24-week regimen of TVR-based triple therapy for Japanese genotype 1 CHC patients. These recommendations are based on baseline factors, including patient’s age, sex, IL28B genotype, core amino acid substitution at position 70, previous treatment history and response, stage of fibrosis, viral load, and baseline hemoglobin level [25, 26].

In addition to the baseline predictive factors, changes in HCV RNA levels after the start of therapy are predictive for treatment outcomes. A rapid virological response (RVR), defined as undetectable serum HCV RNA at week 4 after the start of therapy, and an extended rapid virological response, defined as undetectable serum HCV RNA at both weeks 4 and 12, were also reported as significant predictors of TVR-based treatment outcome [7, 11, 15, 16, 27]. However, the association between SVR and viral kinetics earlier than 4 weeks after initiating TVR-based triple combination therapy remains unclear. RVR was achieved in only approximately 3–11 % of cases receiving PEG-IFN and RBV combination therapy [2, 3, 6, 27, 28]. In contrast, RVR was achieved in approximately 61–84 % of cases receiving TVR-based triple combination therapy [26, 8, 10, 11, 15, 16, 27, 29]. It is therefore important to determine whether viral kinetics earlier than week 4 after the start of therapy is predictive for SVR in TVR-based triple combination therapy?

TVR-based triple combination therapy remarkably improves the SVR rate in CHC patients with the difficult-to-treat HCV genotype 1. However, some patients still fail to achieve SVR. Adverse events occurred more frequently and were more severe in patients treated with TVR-based therapy than in those treated with PEG-IFN and RBV alone [26]. Additionally, TVR-based therapy is expensive. In clinical practice, the determination of predictive factors of successful treatment outcome as early as possible is necessary for preventing unnecessary treatment in addition to physical and economic burden. Thus, in this prospective, multicenter study, we evaluated the clinical relevance of baseline predictors and the reduction in HCV RNA levels at week 1 after starting therapy for predicting SVR in a 24-week regimen of TVR-based triple combination therapy for genotype 1b CHC patients.

Methods

Patients, treatment, and definition of outcomes

Between December 2011 and September 2012, 156 Japanese genotype 1b monoinfected CHC patients were enrolled in this multicenter study at Shinmatsudo Central General Hospital, Kurume University School of Medicine, Kagawa Prefectural Central Hospital, Jikei University School of Medicine Kashiwa Hospital, and Ogaki Municipal Hospital. The inclusion criteria for the study included persistently positive sera for HCV RNA for > 6 months as determined using the quantitative real-time PCR method (COBAS AmpliPrep/COBAS TaqMan HCV test, Roche Diagnostics, Tokyo, Japan), HCV RNA ≥5.0 log10IU/mL in treatment-naïve patients, age of 18–75 years, and body weight >35 kg at the time of entry into the study. Exclusion criteria were: (1) decompensated cirrhosis; (2) positive for hepatitis B surface antigen or antibodies against human immunodeficiency virus; (3) previous or current development of hepatocellular carcinoma; (4) co-existence of other liver diseases, such as autoimmune hepatitis, primary biliary cirrhosis, hemochromatosis, Wilson disease, and alcoholic liver disease; (5) renal disease or creatinine clearance ≤ 50 mL/min at baseline; (6) hemoglobin level < 12 g/dL, white blood cell count < 2000/μL, neutrophil count < 1500/μL, and platelet count < 8.0 × 104/μL at baseline; (7) depression, schizophrenia or its history, or history of suicide attempts, (8) pregnancy in progress or planned for either partner during the study period. For 114 of 156 (73.1 %) patients, liver biopsy was conducted within 12 months of enrollment. The presence or absence of cirrhosis was established according to the Metavir score [30]. For the remaining 42 patients, the presence or absence of cirrhosis was evaluated using ultrasonography and/or computed tomography findings.

Patient profiles are shown in Table 1. In this study, all treatment-experienced patients were treated with PEG-IFN and ribavirin combination therapy. Patients in this study were categorized as relapsers (HCV RNA undetectable at the end of treatment and then positive in follow-up), partial responders (≥2 log10IU/mL reduction in HCV RNA at week 12 but never undetectable), or null responders (< 2 log10IU/mL reduction in HCV RNA at week 12). In this study, partial responders and null responders were analyzed as non-responders.
Table 1

Patient profiles

Number of patients

156

Sex (male/female)

78/78

Age (years)

58.4 ± 10.3

Body weight (kg)

61.8 ± 12.8

Body mass index (kg/m2)

23.7 ± 3.5

Absence or presence of cirrhosis (non-cirrhosis/cirrhosis)

120/36

Response to previous treatment (treatment-naïve/relapsers/partial responders/null responders)

78/50/14/14

rs8099917 (TT/TG/GG)

106/48/2

Core amino acid substitution 70 (wild-type/mutant-type)

97/59

ISDR of NS5A (wild-type/non-wild-type)

138/18

White blood cells (/μL)

4972 ± 1542

Hemoglobin (g/dL)

14.2 ± 1.4

Platelets (×104/μL)

17.1 ± 5.6

Aspartate aminotransferase I (U/L)

54 ± 36

Alanine aminotransferase I (U/L)

60 ± 50

Gamma-glutamyl-transpeptidase I (U/L)

59 ± 68

Albumin (g/dL)

4.2 ± 0.3

Total cholesterol (mg/dL)

173 ± 31

Low-density lipoprotein cholesterol (mg/dL)

103 ± 28

Alpha-fetoprotein (ng/mL)

10.9 ± 20.7

HCV RNA (log10IU/mL)

6.4 ± 0.9

Initial dose of PEG-IFN (μg/kg)

1.5 ± 0.2

Initial dose of ribavirin (mg/kg)

11.2 ± 1.6

Initial daily dose of telaprevir (1500/2250 mg)

84/72

Administration intervals of telaprevir (q8/q12 h)

96/60

Data are expressed as numbers or mean ± standard deviation

ISDR interferon sensitivity-determining region, HCV hepatitis C virus, PEG-IFN PEG-interferon

All patients were treated with PEG-IFN-α-2b, RBV, and TVR triple therapy. TVR (Telavic; Mitsubishi Tanabe Pharma, Osaka, Japan) was administered every 8 h after meals (q8 h) at 500 or 750 mg, or every 12 h after meals (q12 h) at 750 or 1125 mg. The initial daily dose of TVR (1500 or 2250 mg per day) and administration intervals (q8 or q12 h) were determined by each attending physician according to age, sex, body weight, and hemoglobin level. PEG-IFN-α-2b (PEG-Intron, MSD, Tokyo, Japan) was injected subcutaneously at a median dose of 1.5 μg/kg per week. The RBV (Rebetol, MSD, Tokyo, Japan) dose was adjusted by body weight (600 mg for < 60 kg; 800 mg for ≥60 to <80 kg; and 1000 mg for ≥80 kg; in the case of hemoglobin < 13 g/dL at start of therapy, the RBV dose was reduced by 200 mg), based on the guidelines of the Ministry of Health, Labor and Welfare of Japan, and the drug was administered orally after breakfast and dinner. Triple therapy was given for 12 weeks, followed by an additional 12 weeks of PEG-IFN-α-2b and RBV combination therapy (T12PR24). Administration of each drug was appropriately reduced or withdrawn when a serious adverse event was suspected to be developing or if a serious adverse event occurred during the course of treatment. Regardless of adverse events, treatment was stopped for patients who had HCV RNA > 3 log10IU/mL at week 4 or detectable HCV RNA at week 12, or those showing a > 2 log10IU/mL increase in HCV RNA levels from the lowest level during therapy, because of the low likelihood of achieving SVR and the high risk of developing antiviral resistance.

Adherence to PEG-IFN was calculated based on the initial weekly dose, and that to RBV was calculated based on the initial daily dose. Adherence to TVR was defined as 100 % when 2250 mg was given each day for 12 weeks, which is the recommended daily dose.

The virological response was analyzed on an intent-to-treat basis. The successful endpoint of treatment was SVR for patients showing undetectable HCV RNA for 24 weeks after cessation of treatment. Patients were defined as relapse when HCV RNA levels became undetectable until the end of treatment, but became positive during the follow-up period. Patients were defined as at viral breakthrough when HCV RNA became undetectable during the treatment period, but then became positive again before the end of the treatment period. Patients were defined as non-response when HCV RNA was detectable throughout the treatment period. Furthermore, RVR was defined as undetectable HCV RNA at week 4 after starting treatment.

All patients provided written informed consent. This study protocol was prepared following ethics guidelines established in conformity with the 2008 Declaration of Helsinki, and was approved by the Ethics Committee of each participating institution.

Measurement of HCV RNA, and amino acid substitution in the core and NS5A regions of HCV genotype 1b

HCV genotype was determined by direct sequencing followed by phylogenic analysis of the NS5B region [31]. The antiviral effects of the therapy on HCV were assessed by measuring serum HCV RNA levels. In this study, HCV RNA levels were evaluated at baseline; weeks 1, 4, 8, 12, 16, 20, and 24 during treatment; and once every 4 weeks after cessation of treatment. HCV RNA levels were determined using the COBAS AmpliPrep/CABAS TaqMan HCV test (Roche Diagnostics, Tokyo, Japan). The linear dynamic range of the assay was 1.2–7.8 log10IU/mL, and undetectable samples were defined as negative.

Core amino acid substitution at position 70 was determined according to a previously described method [32, 33]. Core amino acid substitution at position 70 was defined as wild-type (arginine) or mutant-type (glutamine or histidine). Additionally, substitutions at amino acids 2290–2248 of the NS5A region [interferon-sensitivity determining region (ISDR)] were determined using a previously described method [34]. Amino acid substitutions in ISDR were defined as wild-type (0 or 1) or non-wild-type (≥2).

Single-nucleotide polymorphism genotyping

Genomic DNA was extracted from whole blood using the MagNA Pure LC and a DNA Isolation Kit (Roche Diagnostics). The genetic polymorphism rs8099917, near the IL28B gene [17, 18], was genotyped by real-time detection PCR using the TaqMan SNP Genotyping Assays and the 7500Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The rs8099917 genotypes were classified into 2 categories, including TT (major genotype) and non-TT genotype (minor genotype: TG or GG).

Statistical analysis

Continuous variables are expressed as the mean and standard deviation. Categorical data were analyzed using the Chi-squared test and Fisher’s exact test, while continuous data were analyzed using the non-parametric Mann–Whitney U test. Univariate and multiple logistic regression analyses were used to identify factors that significantly contributed to SVR. The odds ratios (OR) and 95 % confidence intervals (95 % CI) were also calculated. All P values for statistical tests were 2-tailed, and values of < 0.05 were considered statistically significant. Variables that achieved statistical significance (P < 0.05) according to univariate analysis were entered into multiple logistic regression analyses to identify significant independent predictive factors of SVR.

Receiver-operating characteristics (ROC) analyses were performed to determine cut-off values for sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy for predicting SVR. Statistical analysis was performed using SPSS version 17.0 (IBM-SPSS, Chicago, IL, USA).

Results

Characteristics of patients and treatment outcome

Table 1 summarizes the characteristics of the patients. In total, 78 patients (50.0 %) were treatment-naïve, and 78 patients (50.0 %) were treatment-experienced with PEG-IFN and RBV. The IL28B TT genotype was present in 67.9 % (106 of 156) of the patients. The proportion of patients with cirrhosis was 23.1 % (36 of 156). In total, 72 patients (46.1 %) were treated with TVR at 2250 mg/day, and 84 patients (53.9 %) were treated with TVR at 1500 mg/day. In terms of dosing schedule, 96 patients (61.5 %) were treated q8 h, and 60 patients (38.5 %) were treated q12 h.

Regarding treatment outcomes, 125 patients (80.1 %) achieved SVR; 14 patients (9.0 %) relapsed. 12 patients (7.7 %) showed viral breakthrough, and the remaining five patients (3.2 %) showed non-response. For the IL28B SNP genotypes, among the 106 patients with the TT genotype, 102 (96.2 %) achieved an SVR, and one (0.9 %) relapsed; two (1.9 %) showed viral breakthrough, and one (0.9 %) showed non-response Among the 50 patients with the non-TT genotype, 23 (46.0 %) achieved an SVR; 13 (26.0 %) relapsed. Ten (20.0 %) showed viral breakthrough, and four (8.0 %) showed non-response. Thus, the SVR rate was significantly higher in patients with the TT genotype than in those with the non-TT genotype [102 of 106 patients (96.2 %) vs. 23 of 50 (46.0 %), P < 0.0001] (Fig. 1). According to previous treatment response, among the 78 treatment-naïve patients, 66 (84.6 %) achieved an SVR; five (6.4 %) relapsed. Five (6.4 %) showed viral breakthrough, and two (2.6 %) showed non-response. Among the 50 relapsers, 48 (96.0 %) achieved an SVR; one (2.0 %) relapsed, and one (2.0 %) showed viral breakthrough. Among the 14 partial responders, eight (57.1 %) achieved an SVR; four (28.6 %) relapsed, and two (14.3 %) showed viral breakthrough. Among the 14 null responders, three (21.4 %) achieved an SVR; four (28.6 %) relapsed. Four (28.6 %) showed viral breakthrough, and three (21.4 %) showed non-response. The SVR rate was significantly different across the four categories of previous treatment response (P < 0.0001). In particular, the SVR rate was significantly lower in non-responders than in treatment-naïve patients or relapsers [114 of 128 patients (89.1 %) vs. 11 of 28 patients (39.3 %), P < 0.0001].
https://static-content.springer.com/image/art%3A10.1007%2Fs00535-013-0918-7/MediaObjects/535_2013_918_Fig1_HTML.gif
Fig. 1

Rate of sustained virological response according to the IL28B (rs8099917) genotype. The rate of sustained virological response was significantly higher in patients with the TT genotype than in those with the non-TT genotype (P < 0.0001)

Six patients stopped triple therapy before 12 weeks. The reasons were loss of appetite in three patients, severe anemia in one patient, systemic skin flare in one patient, and viral breakthrough in one patient. Among the six patients, five (83.3 %) with the IL28B TT genotype achieved an SVR, and one (16.7 %) with the non-TT genotype who showed viral breakthrough did not achieve an SVR.

Association between reduced serum HCV RNA levels at week 1 after starting therapy and SVR

ROC curve analysis was performed in 156 patients, to evaluate the association between reduced serum HCV RNA levels at week 1 after starting therapy and SVR. The area under the ROC curve was 0.754, and the best cut-off value was calculated as 4.7 log10IU/mL (Fig. 2). The SVR rate was significantly higher in patients with a reduction of ≥4.7 log10IU/mL at week 1 than in those with a reduction of <4.7 log10IU/mL [65 of 68 patients (95.6 %) with ≥4.7 log10IU/mL vs. 60 of 88 patients (68.2 %) with <4.7 log10IU/mL, P < 0.0001]. All four patients with the TT genotype who failed to show an SVR had a reduction of <4.7 log10IU/mL at week 1 (4.1 log10IU/mL in treatment-naïve patient, 3.8 log10IU/mL in partial responder, 3.7 log10IU/mL in null responder, and 4.6 log10IU/mL in null responder, respectively).
https://static-content.springer.com/image/art%3A10.1007%2Fs00535-013-0918-7/MediaObjects/535_2013_918_Fig2_HTML.gif
Fig. 2

Receiver operating characteristic (ROC) analysis for prediction of a sustained virological response according to the reduction in serum HCV RNA levels at week 1 after the start of therapy. The area under the ROC curve was 0.754

Patients with the IL28B TT genotype presented an extremely high SVR rate. Therefore, the ROC analysis focused on 50 patients with the IL28B non-TT genotype. The area under the ROC curve was 0.777, and the best cut-off value was calculated as 4.7 log10IU/mL, which was similar to the value calculated for all patients (Fig. 3). The SVR rate was significantly higher in patients with a reduction of ≥4.7 log10IU/mL at week 1 than in those with a reduction of <4.7 log10IU/mL [15 of 18 patients (83.3 %) with a reduction of ≥4.7 log10IU/mL vs. 8 of 32 patients (25.0 %) with a reduction of <4.7 log10IU/mL, P = 0.0001].
https://static-content.springer.com/image/art%3A10.1007%2Fs00535-013-0918-7/MediaObjects/535_2013_918_Fig3_HTML.gif
Fig. 3

Receiver operating characteristics (ROC) analysis for prediction of a sustained virological response in the IL28B (rs8099917) non-TT genotype according to the reduction in serum HCV RNA levels at week 1 after the start of therapy. The area under the ROC curve was 0.777

Predictive factors associated with SVR

According to the univariate analysis, the following factors were associated with SVR: treatment-naïve patients or relapsers (P < 0.0001); IL28B TT genotype (P < 0.0001); higher white blood cell count (P = 0.0098), platelet count (P = 0.0299), total cholesterol level (P = 0.0467), and low-density lipoprotein cholesterol level (P = 0.0080); lower gamma glutamyl transpeptidase level (P = 0.0014) and alpha-fetoprotein level (P = 0.0175); core amino acid substitution at position 70 of the wild-type (P = 0.0010); achievement of RVR (P < 0.0001); and reduction of ≥4.7 log10IU/mL in HCV RNA levels at week 1 (P = 0.0003). Multiple logistic regression analysis identified the following four independent factors: IL28B TT genotype (P < 0.0001, OR = 73.65, 95 % CI = 11.28–480.93), reduction of ≥4.7 log10IU/mL in HCV RNA at week 1 (P = 0.0029, OR = 18.99, 95 % CI = 2.74–131.63), achievement of RVR (P = 0.0032, OR = 12.59, 95 % CI = 2.33–69.97), and treatment-naïve patients or relapsers (P = 0.0224, OR = 5.58, 95 % CI = 1.28–24.38) (Table 2).
Table 2

Factors associated with sustained virological response

Variable

Simple

Multiple

OR

95 % CI

P value

OR

95 % CI

P value

Host-related factor

 Age (year)

1.00

0.96–1.04

0.9488

   

 Sex male vs. female

1.23

0.56–2.72

0.6019

   

 Body weight (kg)

0.99

0.97–1.02

0.7195

   

 Body mass index (kg/m2)

0.97

0.87–1.08

0.5494

   

 Cirrhosis absence vs. presence

2.20

0.93–5.18

0.0711

   

 Treatment-naïve or relapsers vs. non-responders

12.58

4.92–32.21

< 0.0001

5.58

1.28–24.38

0.0224

 rs8099917 TT vs. non-TT

29.93

9.54–93.92

< 0.0001

73.65

11.28–480.93

< 0.0001

 White blood cells (/μL)

1.00

1.00–1.00

0.0098

   

 Hemoglobin (g/dL)

1.16

0.87–1.55

0.3196

   

 Platelets (×104/μL)

1.09

1.01–1.18

0.0299

   

 Aspartate aminotransferase I(U/L)

0.99

0.98–1.00

0.1034

   

 Alanine aminotransferase I(U/L)

1.00

0.99–1.00

0.3574

   

 Gamma-glutamyl-transpeptidase I(U/L)

0.99

0.99–1.00

0.0014

   

 Albumin (g/dL)

3.14

0.65–15.22

0.1548

   

 Total cholesterol (mg/dL)

1.01

1.00–1.03

0.0467

   

 Low-density lipoprotein-cholesterol (mg/dL)

1.03

1.01–1.05

0.0080

   

 Alpha-fetoprotein (ng/mL)

0.97

0.95–1.00

0.0175

   

Virus-related factor

 HCV RNA (log10IU/mL)

1.01

0.64–1.60

0.9695

   

 Core amino acid substitution 70 wild-type vs. mutant-type

4.01

1.75–9.17

0.0010

   

 ISDR of NS5A non-wild-type vs. wild type

2.13

0.46–9.79

0.3319

   

Treatment-response factor

 Rapid virological response + vs. −

9.43

3.89–22.87

< 0.0001

12.59

2.33–69.97

0.0032

 Reduction in HCV RNA level at week 1 ≥4.7 log10/mL vs. <4.7 log10IU/mL

10.11

2.92–34.99

0.0003

18.99

2.74–131.63

0.0029

Treatment-related factor

 Administration intervals of telaprevir q8 vs. q12 h

1.20

0.54–2.67

0.6572

   

 Initial daily dose of telaprevir 2250 vs. 1500 mg

1.46

0.65–3.26

0.3545

   

 Duration of therapy (weeks)

0.66

0.92–1.13

1.0226

   

 Adherence of PEG-IFN (%)

1.00

0.98–1.01

0.5762

   

 Adherence of ribavirin (%)

1.00

1.00–1.00

0.8539

   

 Adherence of telaprevir (%)

1.01

0.99–1.03

0.4877

   

HCV hepatitis C virus, ISDR interferon sensitivity-determining region, Peg-IFN PEG-interferon

When analyses focused on patients with the IL28B non-TT genotype alone, previous relapsers (P = 0.0020), higher white blood cell count (P = 0.0255) and platelet count (P = 0.0161), lower body mass index (P = 0.0400), aspartate aminotransferase level (P = 0.0303), alpha-fetoprotein level (P = 0.0304), achievement of RVR (P = 0.0011), and reduction of ≥4.7 log10IU/mL in HCV RNA levels at week 1 (P = 0.0003) were identified as factors associated with SVR by univariate analysis. The multiple logistic regression analysis identified the following three independent factors: a reduction of ≥4.7 log10IU/mL in HCV RNA at week 1 (P = 0.0043, OR = 29.35, 95 % CI = 2.88–299.22), achievement of RVR (P = 0.0156, OR = 17.96, 95 % CI = 1.73–186.57), and previous relapsers (P = 0.0461, OR = 9.18, 95 % CI = 1.04–81.16) (Table 3).
Table 3

Factors associated with sustained virological response in patients with the IL28B non-TT genotype

Variable

Simple

Multiple

OR

95 % CI

P value

OR

95 % CI

P value

Host-related factor

 Age (year)

0.99

0.94–1.05

0.7963

   

 Sex male vs. female

1.56

0.50–4.83

0.4421

   

 Body weight (kg)

0.97

0.93–1.02

0.2324

   

 Body mass index (kg/m2)

0.82

0.69–0.99

0.0400

   

 Cirrhosis absence vs. presence

1.80

0.50–6.43

0.3657

   

 Relapsers vs. treatment-naïve or non-responders

13.64

2.60–71.46

0.0020

9.18

1.04–81.16

0.0461

 White blood cells (/μL)

1.00

1.00–1.00

0.0255

   

 Hemoglobin (g/dL)

1.26

0.87–1.82

0.2145

   

 Platelets (×104/μL)

1.14

1.02–1.26

0.0161

   

 Aspartate aminotransferase I(U/L)

0.97

0.95–1.00

0.0303

   

 Alanine aminotransferase I(U/L)

0.98

0.96–1.00

0.0564

   

 Gamma-glutamyl-transpeptidase I(U/L)

0.99

0.99–1.00

0.1852

   

 Albumin (g/dL)

2.30

0.42–12.72

0.3380

   

 Total cholesterol (mg/dL)

1.00

0.98–1.02

0.9274

   

 Low-density lipoprotein cholesterol (mg/dL)

1.01

0.99–1.03

0.3557

   

 Alpha-fetoprotein (ng/mL)

0.90

0.82–0.99

0.0304

   

Virus-related factor

 HCV RNA (log10IU/mL)

0.67

0.28–1.59

0.3590

   

 Core amino acid substitution 70 wild-type vs. mutant-type

1.56

0.50–4.83

0.4421

   

 ISDR of NS5A non-wild-type vs. wild type

1.87

0.29–12.33

0.5130

   

Treatment-response factor

 Rapid virological response + vs. −

15.27

2.96–78.81

0.0011

17.96

1.73–186.57

0.0156

 Reduction in HCV RNA level at week 1 ≥4.7 log10/mL vs. <4.7 log10IU/mL

15.00

3.43–65.59

0.0003

29.35

2.88–299.22

0.0043

Treatment-related factor

 Administration intervals of telaprevir q8 vs. q12 h

0.60

0.19–1.84

0.3698

   

 Initial daily dose of telaprevir 2250 vs. 1500 mg

0.71

0.21–2.41

0.5781

   

 Duration of therapy (weeks)

1.16

0.93–1.44

0.1973

   

 Adherence of PEG-IFN (%)

1.04

0.99–1.08

0.1084

   

 Adherence of ribavirin (%)

1.01

0.98–1.04

0.4767

   

 Adherence of telaprevir (%)

0.99

0.96–1.03

0.6851

   

HCV hepatitis C virus, ISDR interferon sensitivity-determining region, Peg-IFN PEG-interferon

Combination of the IL28B genotype and reduction in HCV RNA levels at week 1 after the start of therapy to identify patients with a high likelihood of SVR

Figure 4 shows the schematic representation of the process used to identify patients with a high likelihood to achieve SVR by combining the two factors most strongly associated with SVR. Patients with the IL28B TT genotype presented a high SVR rate [102 of 106 patients (96.2 %)], regardless of the reduction in HCV RNA levels at week 1 after the start of therapy. In contrast, patients with the non-TT genotype showed a high SVR rate [15 of 18 patients (83.3 %)] if they presented a reduction of ≥4.7 log10IU/mL in the HCV RNA levels at week 1 after the start of therapy. In contrast, the SVR rate was significantly lower [8 of 32 patients [25.0 %)] when patients did not present a reduction of ≥4.7 log10IU/mL at week 1 (P = 0.0001). In patients with the IL28B non-TT genotype and a reduction of ≥4.7 log10IU/mL in the HCV RNA levels at week 1, the sensitivity, specificity, PPV, NPV, and accuracy for SVR were 62.5, 88.9, 83.3, 75.0, and 78.0 %, respectively. Furthermore, in patients with the non-TT genotype, when both a reduction of ≥4.7 log10IU/mL in the HCV RNA levels at week 1 and RVR were used, the sensitivity, specificity, PPV, NPV, and accuracy for SVR were 60.9, 100, 100, 75.0, and 82.0 %, respectively.
https://static-content.springer.com/image/art%3A10.1007%2Fs00535-013-0918-7/MediaObjects/535_2013_918_Fig4_HTML.gif
Fig. 4

Prediction of a sustained virological response (SVR) by the IL28B (rs8099917) genotype and reduction in HCV RNA level at week 1 after the start of therapy. In patients with the TT genotype, the SVR rate was high (96.2 %), regardless of the reduction in HCV RNA at week 1. In contrast, in patients with the non-TT genotype, the SVR rate was significantly higher in patients with a reduction of ≥4.7 log10IU/mL in the HCV RNA level at week 1 after the start of therapy than in those with a reduction of <4.7 log10IU/mL in the HCV RNA levels at week 1 [15 of 18 patients (83.3 %) vs. 8 of 32 patients (25.0 %), P = 0.0001]

Discussion

Multiple logistic regression analysis revealed that the IL28B genotype was the most significant factor predicting SVR to a 24-week regimen of TVR-based triple combination therapy. The impact of the IL28B genotype on SVR found for this treatment regimen was in agreement with the findings of previous studies in Japan [7, 11, 1416, 24]. In addition, a reduction of ≥4.7 log10IU/mL in the HCV RNA levels at week 1 after the start of therapy was identified as a strong independent on-treatment predictor for SVR in a multiple logistic regression analysis.

The reduction in HCV RNA levels at week 1 was particularly relevant in patients with the IL28B non-TT genotype. Whereas patients with the IL28B TT genotype showed a high SVR rate regardless of the on-treatment response of HCV RNA, a significant difference in SVR rate was observed based on the reduction in HCV RNA levels at week 1 in patients with the unfavorable non-TT genotype. In this patient subpopulation, the reduction in HCV RNA level at week 1 was the factor most strongly associated with SVR, and this finding is of clinical value to identify patients with a low likelihood of achieving SVR as early as possible. Furusyo et al. [11] previously reported that the serum HCV RNA levels at day 3 presented a significant difference between SVR and non-SVR patients. The ability of the very early viral response to predict SVR shown by both Furusyo et al. and our study may be explained by the strong antiviral effect of TVR. However, Furusyo et al. did not enter serum HCV RNA levels at day 3 into a multiple logistic regression analysis to identify significant independent predictors of SVR. Therefore, in that study, it was not clear whether the serum HCV RNA level at day 3 was an independent factor of SVR when including host-related, virus-related, and on-treatment factors. In the present study, the median serum HCV levels at week 1 was significantly lower for SVR patients (1.9 log10IU/mL) than for non-SVR patients (2.2 log10IU/mL) (P = 0.0136, data not shown). In the present study, the reduction in HCV RNA levels at week 1 after the start of therapy was an independent predictive factor for SVR. This reduction in HCV RNA level at week 1 may represent early viral kinetics closely correlated with the antiviral effect. The predictive ability of the reduction in HCV RNA level at day 3 and week 1 after the start of therapy should therefore be compared based on the IL28B genotype.

This study is the first report to demonstrate that a reduction of ≥4.7 log10IU/mL in the HCV RNA level at week 1 is a useful on-treatment predictive factor associated with SVR to a 24-week TVR-based triple combination therapy in clinical practice, especially in patients with the IL28B non-TT genotype. In ‘real-world’ clinical practice in some cases, it may be impossible to differentiate between previous null and partial responders because of the absence of relevant historical data from medical records. Therefore, for these treatment-experienced patients, IL28B genotyping may have clinical utility, as it may serve as a pretreatment marker for interferon responsiveness to guide patients and physicians. In patients with the IL28B non-TT genotype, both a reduction of <4.7 log10IU/mL in the HCV RNA levels at week 1 and positivity for HCV RNA at week 4 (non-RVR) indicated a high likelihood of treatment failure. Hence, these patients should not undergo TVR-based triple combination therapy to avoid unnecessary treatment. This study identified that measurement of the HCV RNA level not only at week 4, but also at week 1, provides important information for predicting SVR, particularly in patients with the IL28B non-TT genotype.

There were some limitations to this study. First, the number of patients was too low to conclusively identify factors contributing to SVR. In particular, the number of non-responders was very small. Second, TVR-resistant variants were not analyzed. Resistant variants have been reported to occur in 56 % of HCV genotype 1b patients who did not achieve SVR [35]. Therefore, resistance variants should be identified in patients with treatment failure. Third, this study regimen was limited to T12PR24. Only a 24-week TVR-based triple combination therapy (triple therapy for 12 weeks followed by an additional 12 weeks of PEG-IFN and RBV) is allowed by the Japanese National Insurance System. In the US, Canada, and EU, triple combination therapy is administered for either 12 or 36 additional weeks after PEG-IFN and RBV, according to the response-guided regimen based on the early viral response in each category, i.e., treatment-naïve patients and previous relapsers or partial responders and null responders.

Recently, the second-generation direct-acting antiviral agent simeprevir (SMV), which is once-daily oral NS3/4A protease inhibitor, was approved in September 2013 in Japan. Hayashi et al. [36] reported a Japanese phase II study. In treatment-naïve patients, the SVR rate was 77–92 % by triple combination therapy with SMV, PEG-IFN-α-2a and RBV. During the first 3–7 days of SMV-based therapy, an initial rapid reduction in HCV RNA was evident. Mean reduction in HCV RNA at week 1 in our study in TVR-based therapy was 4.5 log10IU/mL (data not shown). Mean reduction in HCV RNA at week 1 was not shown with the numerical value in this SMV-based therapy, but that seems to be similar to our TVR-based therapy. However, in this study, the IL28B genotypes were not investigated. Therefore, in clinical practice, from now on, prospective studies should be necessary to confirm whether the reduction in HCV RNA at week 1 is predictive for SVR in SMV-based therapy based on the IL28B genotype as well as in TVR-based therapy.

In conclusion, this prospective, multicenter study of a 24-week TVR-based triple combination therapy for Japanese genotype 1b CHC patients showed that the IL28B SNP genotype is the most important baseline factor for predicting SVR, and a reduction of ≥4.7 log10IU/mL in the HCV RNA levels at week 1, i.e., viral kinetics earlier than week 4, could be a useful on-treatment predictor of SVR, especially in patients with the IL28B non-TT genotype. Further large-scale prospective studies including SMV-based triple combination therapy are necessary to confirm these findings and develop the individual tailoring and optimization of therapeutics.

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© Springer Japan 2013