Lipids

, Volume 47, Issue 11, pp 1053–1062

High Serum Palmitic Acid is Associated with Low Antiviral Effects of Interferon-Based Therapy for Hepatitis C Virus

Authors

  • Teruki Miyake
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Masashi Hirooka
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Yoshio Tokumoto
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Takao Watanabe
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Shinya Furukawa
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Teruhisa Ueda
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Shin Yamamoto
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Teru Kumagi
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Hiroaki Miyaoka
    • Internal MedicineSaiseikai Matsuyama Hospital
  • Masanori Abe
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Bunzo Matsuura
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
  • Morikazu Onji
    • Department of Gastroenterology and MetabologyEhime University Graduate School of Medicine
Original Article

DOI: 10.1007/s11745-012-3716-8

Cite this article as:
Miyake, T., Hiasa, Y., Hirooka, M. et al. Lipids (2012) 47: 1053. doi:10.1007/s11745-012-3716-8
  • 237 Views

Abstract

Hepatitis C virus (HCV) infection alters fatty acid synthesis and metabolism in association with HCV replication. The present study examined the effect of serum fatty acid composition on interferon (IFN)-based therapy. Fifty-five patients with HCV were enrolled and received IFN-based therapy. Patient characteristics, laboratory data (including fatty acids), and viral factors that could be associated with the anti-HCV effects of IFN-based therapy were evaluated. The effects of individual fatty acids on viral replication and IFN-based therapy were also examined in an in-vitro system. Multivariate logistic regression analysis showed that the level of serum palmitic acid before treatment and HCV genotype were significant predictors for rapid virological response (RVR), early virological response (EVR), and sustained virological response (SVR). High levels of palmitic acid inhibited the anti-HCV effects of IFN-based therapy. HCV replication assays confirmed the inhibitory effects of palmitic acid on anti-HCV therapy. The concentration of serum palmitic acid is an independent predictive factor for RVR, EVR, and SVR in IFN-based antiviral therapy. These results suggest that the effect of IFN-based antiviral therapy in patients with HCV infection might be enhanced by treatment that modulates palmitic acid levels.

Keywords

Fatty acidHepatitis C virusInterferonPalmitic acidVirological response

Abbreviations

AUC

Area under curve

BMI

Body mass index

DMEM

Dulbecco’s modified Eagle’s medium

EVR

Early virological response

GAPDH

Glyceraldehyde-3-phosphate dehydrogenase

HCC

Hepatocellular carcinoma

HBc

Hepatitis B core

HBs-Ag

Hepatitis B surface antigen

HBV

Hepatitis B virus

HCV

Hepatitis C virus

IFN

Interferon

IL-28B

Interleukin-28B

ISDR

Interferon sensitivity-determining region

NPV

Negative predictive value

NS5

Nonstructural-5

PEG-IFN

Pegylated interferon

PPV

Positive predictive value

RBV

Ribavirin

ROC

Receiver operating characteristics

RVR

Rapid virological response

SVR

Sustained virological response

PCR

Polymerase chain reaction

Introduction

Hepatitis C virus (HCV) infection is common worldwide, and more than 80 % of patients develop chronic infection. Of those with chronic infection, 20–30 % develop cirrhosis and hepatocellular carcinoma (HCC) [1]. Interferon (IFN)-based treatment regimens have been widely used, but these treatment regimens have side effects, require long-term therapy, and are expensive. Estimation of the effectiveness of IFN-based therapy prior to treatment would be beneficial.

The velocity of decrease in viral load during IFN-based therapy is a good indicator for the prediction of sustained virological response (SVR). High SVR rates are predicted by rapid virological response (RVR) and early virological response (EVR) [2, 3]. Although mechanisms of treatment failure are poorly understood, previous reports have proposed IFN-stimulated genes and the inability to develop effective anti-HCV immunity as possible explanations [4].

Recently, fatty acids have been implicated in the pathogenesis of several diseases associated with metabolic disorders (such as obesity, diabetes and cardiovascular disease) [5, 6] and in immunological response [7]. In liver diseases, especially in non-alcoholic steatohepatitis, the effect of impaired peroxisomal polyunsaturated fatty acid metabolism and nonenzymatic oxidation on fatty acid constitution is associated with disease progression [8]. It has been reported that HCV core protein has effects on fatty acid synthesis, and that fatty droplets in the liver are related to development of disease [9, 10]. Although there have been reports about the role of fatty acids in liver in patients with HCV [9, 10], the relationship between serum fatty acids and efficacy of IFN-based antiviral therapy against HCV remains controversial.

The aims of the present study were to evaluate whether the composition of serum fatty acids could predict RVR, EVR, or SVR from IFN-based therapy in patients with HCV. Data from HCV patients were collected and in-vitro assays were performed using HCV-replication cell culture systems.

Patients and Methods

Patients

Consecutive patients with HCV treated with IFN-based therapy at Ehime University Hospital were enrolled prospectively from December 2008 to December 2010. Moreover, 10 healthy volunteers (age range 26–70 years) were enrolled in this study as healthy subjects. This study was carried out in accordance with the Declaration of Helsinki, and the institutional review board of Ehime University Hospital approved this study (Approval # 0710004). Written informed consent was provided by study participants.

Patients with chronic HCV infection, with creatinine clearance >50 mL/min, and who had not been previously treated with antiviral or immunosuppressive agents within the 3 months preceding enrollment were included. Patients with other liver disease such as autoimmune hepatitis, primary biliary cirrhosis, hepatitis B virus (HBV) infection, or HCC; co-infection with human immunodeficiency virus; poorly controlled cardiovascular, hematologic or pulmonary disease; pregnancy; autoimmune disease; severe depression or other psychiatric disorders; and/or active substance abuse were excluded. To exclude HBV infection, hepatitis B surface antigen (HBs-Ag) and anti-hepatitis B core (HBc) antibody were checked. Patients with positive HBs-Ag or a high titer of anti-HBc antibody were excluded.

Interferon and Ribavirin Combination Therapy

Patients received one of four treatment regimens: [1] pegylated interferon (PEG-IFN) α-2b 1.0 μg/kg/week or 1.5 μg/kg/week subcutaneously in combination with oral ribavirin (RBV) dosed by body weight (40–65 kg, 800 mg/day; >65–85 kg, 1,000 mg/day; >85–105 kg, 1,200 mg/day; >105–125 kg, 1,400 mg/day), [2] PEG-IFN α-2a 180 μg/week subcutaneously plus oral RBV dosed as above, [3] PEG-IFN α-2a 180 μg/week subcutaneously, or [4] IFN-β 600 million IU/day intravenously plus oral RBV dosed as above. In order to identify the levels of fatty acids that could have an important role in the response to treatment with IFN, all patients who had been treated with IFN were enrolled. After informing potential subjects about the costs, estimated adverse events, and effects of each treatment protocol according the Japanese guidelines for anti-HCV treatment [11], the required treatment regimen was chosen and treatment was started.

Laboratory Assessment

Patients’ serum samples were collected around 6 a.m. after fasting on day 2 of the study before IFN-based treatment. Additionally, fasting serum was collected at the end of IFN-based treatment. Serum samples were frozen and stored at −80 °C within 4 h of collection and then thawed at the time of measurement. Fatty acid concentrations in total serum lipids was measured with liquid chromatography (SRL Co. Ltd., Tokyo, Japan). Subjects were diagnosed as having dyslipidemia if they had TC ≥220 mg/dL, and/or HDL-c ≤40 mg/dL, and/or TG ≥150 mg/dL [12], and/or taking lipid-lowering agents.

The HCV genotype was determined by the polymerase chain reaction (PCR) using a mixed primer set derived from nucleotide sequences from the nonstructural-5 (NS5) region (SRL Co. Ltd.) [13]. HCV RNA was measured quantitatively before and during therapy using PCR (Cobas Amplicor HCV monitor v 2.0 using the tenfold dilution method, Roche Diagnostics, Mannheim, Germany). The lower level of detection for this assay was less than 1.2 log10 IU/mL. Undetectable serum HCV RNA on testing was considered a negative test. RVR was defined as undetectable serum HCV RNA within 4 weeks from the start of the treatment. EVR was defined as undetectable serum HCV RNA within 12 weeks from the start of the treatment. SVR was defined as undetectable serum HCV RNA within 24 weeks after the end of treatment.

The HCV genotype recovered from patients was determined using the Illumina Human610-quad BeadChip (Illumina, San Diego, CA, USA) as previously described [14, 15]. Amino acid substitutions of aa70 or aa91 in the core region of HCV genotype 1b were evaluated by agarose gel electrophoresis using mutation-specific primers for wild-type (aa70: arginine, aa91: leucine) and mutant (aa70: glutamine/histidine, aa91: methionine) viruses (SRL Co. Ltd.) [16]. In this study, the pattern of arginine (wild-type) at aa70 and leucine (wild-type) at aa91 was evaluated as double wild-type, while the other patterns were non-double wild-type. The nucleotide sequence of the interferon sensitivity-determining region (ISDR) in the HCV NS5A region was determined by direct sequencing through PCR-amplified materials [17]. Wild-type ISDR was defined as having no amino acid substitutions based on the HCV-J strain of genotype 1b.

Assay for Detecting Single Nucleotide Polymorphisms of Interleukin-28B (IL-28B)

Two single nucleotide polymorphisms of interleukin-28B (rs8099917 [14] and rs12979860 [15]) were examined using the TaqMan assay. The sequence of the probe and primers for the TaqMan assay for detecting rs8099917 was provided by Dr. Yasuhito Tanaka (Department of Virology and Liver Unit, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan), and rs12979860 was provided by Dr. David B. Goldstein (Center for Human Genome Variation, Duke University, Durham, NC, USA). Patient DNA was isolated from blood samples. For rs8099917 [14], homozygotes (T/T) were defined as having the IL-28B major allele, and heterozygotes (T/G) or homozygotes (G/G) were defined as having the minor allele. For rs12979860 [25], homozygotes (C/C) were defined as having the major allele, and heterozygotes (T/C) or homozygotes (T/T) were defined as having the minor allele.

Preparation of the In-Vitro Replication System

The human hepatoma cell lines Huh7 (Japanese Collection of Research Bioresources, Osaka, Japan) and Huh7.5.1 (provided by Dr. Francis V. Chisari; Department of Immunology and Microbial Science, The Scripps Research Institute, La Jolla, CA, USA) were cultured with Dulbecco’s modified Eagle’s medium (DMEM) (Sigma Chemical, St. Louis, MO, USA) containing 10 % fetal bovine serum (Filtron PTY LTD, Brooklyn, Australia).

For the in-vitro assay of HCV genotype 1, the plasmid-based binary HCV replication system was used [18, 19], in which the plasmid contained the infectious full-length genotype 1a cDNA sequence corresponding to the H77 prototype strain with the T7 promoter sequence (pT7-flHCV-Rz, provided by Dr. Raymond T. Chung (Gastrointestinal Unit, Massachusetts General Hospital, Boston, MA, USA). pT7-flHCV-Rz cells were transfected to Huh7 cells by using Lipofectamine (Invitrogen, Carlsbad, CA, USA). Subsequently, the T7 polymerase was delivered by using a replication-defective adenovirus vector (Ad-T7pol) at a multiplicity of infection of 10.

For the in-vitro assay of HCV genotype 2, the HCV replication system pJFH1-full that encodes HCV genotype 2a JFH1 sequence was used, which was provided by Dr. Takaji Wakita (Department of Virology II, National Institute of Infectious Diseases, Tokyo, Japan) [20]. HCV RNA was synthesized using the Megascript T7 kit (Ambion, Austin, TX, USA), with the linearized pJFH1-full as template. After DNase I (Ambion) treatment, the transcribed HCV RNA was purified using ISOGEN-LS (Nippon Gene, Tokyo, Japan). For RNA transfection, Huh 7.5.1 cells were resuspended in Opti-MEM I (Invitrogen) containing 10 μg of HCV RNA and subjected to an electric pulse (960 μF and 260 V) using the Gene Pulser II apparatus (Bio-Rad, Richmond, CA, USA). After electroporation, the cell suspension was cultured under normal conditions with DMEM.

Evaluation of Effect of Palmitic Acid In Vitro

For the in-vitro assay, palmitic acid (Sigma Chemical), myristic acid (Sigma Chemical), stearic acid (Sigma Chemical), and oleic acid (Sigma Chemical), were solubilized in ethanol with albumin as a stock solution of 20 mM and stored at −20 °C, as described previously [21, 22]. These fatty acid-albumin complex solutions were freshly prepared before each experiment. Subsequently, preliminary experiments were performed using 10–500 μM solutions of fatty acids in order to assess the concentrations of fatty acids that would not compromise cell viability. Cell viability was not compromised when 10–100 μM fatty acids were used in the MTS assay (cytotoxicity assay using 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, CellTiter 96 Aqueous One Solution cell proliferation assay®, Promega, Madison, WI) [18]; thus, 100 μM fatty acids were used for culture experiments. Fatty acids were added after the preparation of H77 and JFH1 HCV replication cells, and then IFN-α2b (100 IU/mL) and RBV (50 μM) were added to the culture medium of those cells. In the control samples, solubilized solutions were added without fatty acids.

Quantitative Real-Time Reverse-Transcription PCR

Cellular mRNA was extracted by TRIzol (Invitrogen), and levels of HCV replication were quantified by real-time reverse transcription-PCR with SYBR green I dye (Roche Diagnostics) and primers encoded for the 5′UTR of HCV using LightCycler technology (Roche Diagnostics) as described previously [18, 19]. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Roche Diagnostics) was detected using primer sets under the recommended conditions. Data are expressed as copy numbers of HCV RNA or cellular mRNA per molecule of GAPDH mRNA.

Statistical Analysis

Data are expressed as means ± standard deviations or as means ± standard errors of mean. The Wilcoxon test was used to analyze continuous variables. The Chi-square test was used for analysis of categorical data. Group comparisons involving more than two independent groups were performed using the Kruskal–Wallis test. Multivariate analysis was performed using a logistic regression model with stepwise method. The relationships between palmitic acid and the other parameters were analyzed with the Pearson product-moment correlation coefficient. Each cutoff point for continuous variables was decided by receiver operating characteristics (ROC) curve analysis. The RVR, EVR, and SVR rates of patients with low levels of palmitic acid were defined as positive predictive value (PPV) in the prediction of RVR, EVR, and SVR. Non-RVR, non-EVR, and non-SVR rates in patients with high levels of palmitic acid were defined as the negative predictive value (NPV) for prediction of non-RVR, non-EVR, and non-SVR. A P value of less than 0.05 was considered statistically significant. Statistical analyses were performed using JMP version 9 software (SAS Institute Japan, Tokyo, Japan).

Results

Patient Characteristics

A total of 55 patients with HCV were included. There were 27 men and 28 women. The mean age was 53.59 ± 11.15 years (range 22–69 years). The baseline characteristics of the study population are shown in Supplemental Table 1. Thirty-seven patients were infected with genotype 1 (all genotype 1b) and 18 patients with genotype 2 (14 with genotype 2a and four with genotype 2b). There was no significant difference between two groups in serum lipid composition (Table 1).
Table 1

Total serum lipid composition (μg/mL) in 55 patients infected with HCV

Genotype

All (n = 55)

1b (n = 37)

2a or b (n = 18)

P value (1b vs. 2a, b)

Saturated fatty acid

 Lauric acid (12:0)

1.31 ± 1.1

1.2 ± 0.81

1.53 ± 1.54

0.73

 Myristic acid (14:0)

24.68 ± 17.47

21.91 ± 8.38

30.37 ± 27.75

0.37

 Palmitic acid (16:0)

614.44 ± 138.03

598.02 ± 133.65

648.19 ± 144.58

0.35

 Stearic acid (18:0)

186.63 ± 43.27

180.76 ± 40.53

198.71 ± 47.32

0.25

 Arachidic acid (20:0)

5.98 ± 1.3

5.84 ± 1.34

6.27 ± 1.19

0.25

 Behenic acid (22:0)

15.75 ± 3.97

15.31 ± 3.81

16.63 ± 4.25

0.29

 Lignoceric acid (24:0)

14.11 ± 3.23

13.88 ± 3.13

14.59 ± 3.47

0.62

Monounsaturated fatty acid

 Palmitoleic acid (16:1n7)

66.96 ± 26.69

63.96 ± 27.34

73.14 ± 24.92

0.23

 Oleic acid (18:1n9)

537.07 ± 151.73

510.11 ± 126.91

592.48 ± 184.90

0.21

 Eicosenoic acid (20:1n9)

4.65 ± 1.59

4.42 ± 1.28

5.11 ± 2.04

0.22

 Erucic acid (22:1n9)

1.39 ± 0.51

1.35 ± 0.42

1.47 ± 0.66

0.31

 Nervonic acid (24:1n9)

33.25 ± 5.8

33.34 ± 6.04

33.07 ± 5.44

0.94

Polyunsaturated fatty acid

 Linoleic acid (18:2n6)

693.88 ± 151.44

680.31 ± 155.32

721.76 ± 143.32

0.23

 γ-Linolenic acid (18:3n6)

7.57 ± 3.84

7.15 ± 3.62

8.44 ± 4.24

0.3

 α-Linolenic acid (18:3n3)

19.31 ± 8.26

18.14 ± 7.12

21.73 ± 10.01

0.27

 Eicosadienoic acid (20:2n6)

5.72 ± 1.51

5.72 ± 1.56

5.71 ± 1.45

0.91

 Mead acid (20:3n9)

2.18 ± 1.19

2.25 ± 1.33

2.02 ± 0.84

0.80

 Dihomo-γ-linolenic acid (20:3n6)

35.68 ± 11.75

35.14 ± 12.76

36.8 ± 9.57

0.6

 Arachidonic acid (20:4n6)

140.91 ± 36.63

139.61 ± 39.44

143.58 ± 30.91

0.61

 Eicosapentaenoic acid (20:5n3)

45.5 ± 26.19

44.28 ± 24.04

48 ± 30.74

0.67

 Adrenic acid (22:4n6)

4.59 ± 1.56

4.53 ± 1.69

4.71 ± 1.27

0.50

 Docosapentaenoic acid (22:5n3)

17.63 ± 6.64

18.08 ± 5.72

18.77 ± 8.30

0.75

 Docosahexaenoic acid (22:6n3)

123.37 ± 44

120.5 ± 39.73

129.28 ± 52.46

0.7

Values are expressed as means ± standard deviation

P values were determined by the Wilcoxon test

Response to Treatment

Among patients with genotype 1, 22 % (8/37) of patients received PEG-IFN α-2a plus RBV, 16 % (6/37) received PEG-IFN α-2a, 51 % (19/37) received PEG-IFN α-2b plus RBV, and 11 % (4/37) received IFN-β plus RBV. Among patients with genotype 2, 6 % (1/18) of patients received PEG-IFN α-2a plus RBV, 22 % (4/18) received PEG-IFN α-2a, 67 % (12/18) received PEG-IFN α-2b plus RBV, and 6 % (1/18) received IFN-β plus RBV.

RVR was achieved in 35 % (19/55) of patients overall. Twenty-seven percent (10/37) of patients with HCV genotype 1, and 50 % (9/18) with genotype 2 achieved RVR. EVR occurred in 64 % (35/55) of patients. EVR was achieved in 51 % (19/37) of patients with HCV genotype 1, and in 89 % (16/18) of patients with genotype 2.

For assessment of SVR, four patients dropped out of the treatment because of depression (2/4), general fatigue (3/4), and retinopathy (1/4). SVR was achieved in 69 % (35/51) of patients overall, in 56 % (19/34) of patients with HCV genotype 1, and in 94 % (16/17) of patients with genotype 2 evaluated by the above described protocol.

Predictors of Virological Response

Univariate analysis was performed for factors associated with RVR, EVR, and SVR (Supplemental Table 2). For RVR, a low serum triglyceride level was identified as a contributing factor. For EVR, low body mass index (BMI) and HCV genotype 2 were identified as contributing factors. For SVR, low BMI, total cholesterol, and HCV genotype 2 were identified as contributing factors. Further analysis was performed to examine the relationship of fatty acid levels to RVR, EVR, and SVR (Table 2). Table 2 shows the fatty acids significantly associated with virological response to IFN-based therapy by univariate analysis (P < 0.05). Other than the listed fatty acids, lauric acid, arachidic acid, behenic acid, lignoceric acid, myristoleic acid, eicosenoic acid, erucic acid, linoleic acid, γ-linolenic acid, arachidonic acid, and eicosapentaenoic acid were evaluated; however, those fatty acids were not significantly associated with any treatment response (RVR, EVR, or SVR). For RVR, low levels of myristic acid, palmitic acid, stearic acid, oleic acid, α-linolenic acid, eicosadienoic acid, adrenic acid, docosapentaenoic acid, and docosahexaenoic acid were identified as significant contributing factors. For EVR, palmitic acid and nervonic acid were identified as significant contributing factors. For SVR, myristic acid, palmitic acid, palmitoleic acid, nervonic acid, mead acid, dihomo-γ-linoleic acid, and adrenic acid were identified as significant contributing factors. Only a low level of palmitic acid was found to contribute significantly to all of RVR, EVR, and SVR. In Table 2, the odds ratio for each fatty acid was near 1.0, because the range of palmitic acid was wide (from 370.8 to 955.9) compared to the value of the treatment effect of RVR, EVR, and SVR (from 0 to 1).
Table 2

Fatty acids (μg/mL) associated with virological response to interferon-based therapy identified by univariate analysis

 

RVR (n = 55)

EVR (n = 55)

SVR (n = 51)

OR (95 % CI)

P value

OR (95 % CI)

P value

OR (95 % CI)

P value

Saturated fatty acid

 Myristic acid (14:0)

0.88 (0.79–0.96)

0.01

0.96 (0.902–1.002)

0.16

0.93 (0.86–0.99)

0.04

 Palmitic acid (16:0)

0.99 (0.983–0.996)

0.003

0.9956 (0.9909–0.9998)

0.049

0.995 (0.9899–0.9995)

0.04

 Stearic acid (18:0)

0.98 (0.96–0.99)

0.01

0.991 (0.978–1.004)

0.19

0.988 (0.973–1.002)

0.1

Monounsaturated fatty acid

 Palmitoleic acid (16:1n7)

0.98 (0.95–0.99998)

0.07

0.99 (0.97–1.01)

0.36

0.976 (0.953–0.998)

0.04

 Oleic acid (18:1n9)

0.993 (0.987–0.998)

0.004

0.9992 (0.9955–1.003)

0.68

0.998 (0.994–1.002)

0.38

 Nervonic acid (24:1n9)

0.96 (0.86–1.06)

0.41

0.9 (0.8–0.99)

0.04

0.85 (0.75–0.95)

0.01

Polyunsaturated fatty acid

 α-Linolenic acid (18:3n3)

0.86 (0.74–0.96)

0.02

0.98 (0.92–1.05)

0.6

0.97 (0.89–1.04)

0.38

 Eicosadienoic acid (20:2n6)

0.47 (0.25–0.78)

0.01

0.76 (0.5–1.09)

0.15

0.71 (0.45–1.05)

0.11

 Mead acid (20:3n9)

0.68 (0.37–1.13)

0.17

0.73 (0.44–1.17)

0.2

0.51 (0.27–0.86)

0.02

 Dihomo-γ-linolenic acid (20:3n6)

0.949 (0.898–0.998)

0.05

0.957 (0.908–1.004)

0.08

0.92 (0.86–0.97)

0.01

 Adrenic acid (22:4n6)

0.62 (0.38–0.93)

0.03

0.84 (0.58–1.2)

0.33

0.67 (0.43–0.98)

0.048

 Docosapentaenoic acid (22:5n3)

0.74 (0.59–0.88)

0.003

0.96 (0.88–1.04)

0.35

0.93 (0.82–1.03)

0.18

 Docosahexaenoic acid (22:6n3)

0.98 (0.95–0.99)

0.03

0.988 (0.972–1.001)

0.1

0.98 (0.959–0.997)

0.02

Bold values are statistically significant (P < 0.05)

RVR rapid virological response, EVR early virological response, SVR sustained virological response, OR odds ratio, CI confidence interval

Multivariate logistic regression analysis was conducted in order to determine independent predictive variables associated with virological response. With regard to fatty acids, palmitic acid was selected for multivariate analysis because it was a significant factor for RVR, EVR, and SVR by univariate analysis. Palmitic acid concentration was found to be significantly correlated with total cholesterol, triglycerides, and 12 fatty acids by the Pearson product-moment correlation in patients with HCV (Supplemental Table 3). Moreover, the serum levels of palmitic acid were not different statistically by the regimen of IFN based treatment (554.2 ± 96.4 μg/mL in PEG-IFNα-2a + RBV, 556.4 ± 95.8 μg/mL in PEG-IFNα-2a, 643.6 ± 140.7 μg/mL in PEG-IFNα-2b + RBV, and 658.3 ± 209.4 μg/mL in IFNβ + RBV).

For additional factors for multivariate analysis, seven factors were selected that had been identified in previous reports: age, gender, BMI, serum alanine aminotransferase, genotype, viral load, and a history of IFN therapy [2326]. As a result of stepwise multivariate analysis, only low levels of palmitic acid and HCV genotype 2 were found to be significant contributing factors for RVR, EVR, and SVR (Table 3).
Table 3

Factors associated with virological response to interferon-based therapy identified by multivariate analysis

Factor

Category

RVR (n = 55)

EVR (n = 55)

SVR (n = 51)

OR (95 % CI)

P value

OR (95 % CI)

P value

OR (95 % CI)

P value

Genotype

1. 1b

1

 

1

 

1

 

2. 2a, b

0.38 (0.16–0.8)

0.01

0.24 (0.008–0.57)

0.005

0.993 (0.987–0.998)

0.01

Palmitic acid (16:0)

 

0.988 (0.979–0.994)

0.001

0.993 (0.987–0.998)

0.01

0.993 (0.987–0.998)

0.01

Only variables that achieved statistical significance (P < 0.05) on multivariate logistic regression are shown

RVR rapid virological response, EVR early virological response, SVR sustained virological response, OR odds ratio, CI confidence interval

Evaluation of Palmitic Acid Level as a Predictor of Virological Response

ROC curves were constructed and areas under curves (AUC) were calculated (Table 4). A graph of the AUC for RVR, EVR, and SVR is shown in Supplemental Fig. 1. As is seen in Table 4, AUC, cutoff value, sensitivity, specificity, PPV, NPV, and diagnostic accuracy of prediction in RVR were 0.79, 569.1 μg/mL, 73.68, 77.78, 63.64, 84.85, and 76.36 %, respectively; in EVR were 0.7, 586.7 μg/mL, 60, 80, 84, 53.33, and 67.27 %, respectively; and in SVR were 0.7, 587.7 μg/mL, 63.64, 77.78, 84, 53.84, and 68.63 %, respectively. Low levels of palmitic acid in RVR were associated with high specificity and NPV, while low levels of palmitic acid in EVR and SVR were associated high specificity and PPV.
Table 4

Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy of palmitic acid for prediction of virological response to interferon-based therapy

 

AUC (95 % CI)

Cutoff value (μg/mL)

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

Diagnostic accuracy (%)

RVR (n = 55)

0.79 (0.66–0.89)

569.1

73.68

77.78

63.64

84.85

76.36

EVR (n = 55)

0.7 (0.56–0.82)

586.7

60

80

84

53.33

67.27

SVR (n = 51)

0.7 (0.56–0.82)

587.7

63.64

77.78

84

53.84

68.63

Each value was determined by receiver operating characteristic curve analysis

AUC area under curve, RVR rapid virological response, EVR early virological response, SVR sustained virological response

Serum levels of palmitic acid in patients with HCV were significantly higher than those in healthy subjects (614.44 ± 138.03 vs. 480.67 ± 117.53 μg/mL, respectively, P = 0.01). Ninety-one percent (50/55) of patients with HCV had higher levels of palmitic acid compared to the mean level of palmitic acid in healthy subjects. Additionally, serum levels of palmitic acid were evaluated between patients with SVR and patients without SVR. The levels of palmitic acid were significantly lower in patients with SVR compared to patients without SVR (554.3 ± 138.6 vs. 668.5 ± 183 μg/mL, respectively, P = 0.01).

Genetic Analysis

The relationship between palmitic acid and IL-28B polymorphism was investigated in the 37 patients with genotype 1 HCV infection. The level of palmitic acid was compared between major and minor polymorphisms. There were no significant differences between the two groups (Table 5). Amino acid substitution in the core region of HCV and mutated nucleotide sequence of ISDR were not associated with serum levels of palmitic acid (Table 5).
Table 5

Palmitic acid concentration at baseline in 37 patients infected with genotype 1 HCV

 

P value

IL28B polymorphisma

 rs8099917

Major (n = 30)

Minor (n = 7)

0.32

580.8 ± 123.42

672.24 ± 210.48

 rs12979860

Major (n = 29)

Minor (n = 8)

0.14

575.54 ± 104.62

679.51 ± 195.95

Amino acid substitutions in the core regionb

 aa 70

Wild (n = 28)

Non-wild (n = 9)

0.47

588.21 ± 123.42

628.54 ± 166.08

 aa 91

Wild (n = 21)

Non-wild (n = 16)

0.24

571.56 ± 112.74

632.74 ± 153.80

 aa 70 and aa 91

Double-wild (n = 16)

Non-double-wild (n = 21)

0.154

561.71 ± 109.88

625.68 ± 145.73

Nucleotide sequence of ISDRc

 

0–1 (n = 15)

≥2 (n = 21)

0.14

615.12 ± 115.07

564.77 ± 170.04

Data are given as means ± standard deviations

HCV hepatitis C virus, IL-28B interleukin-28B, ISDR interferon sensitivity-determining region, PCR polymerase chain reaction

P value was determined by the Wilcoxon test

aFor IL-28B, the major allele of rs8099917 was defined as TT, and the minor allele was defined as T/G or G/G. The major allele of rs12979860 was defined as C/C, and the minor allele was defined as T/C or T/T

bAmino acid substitutions were evaluated in pretreatment serum by PCR with mutation-specific primers. Wild-type at aa 70 and wild-type at aa 91 were evaluated as double-wild-type, while the other patterns were considered non-double-wild-type

cIn ISDR, 0–1 was defined as having no amino acid substitutions or one substitution, ≥2 was defined as containing two or more amino acid substitutions

In patients with HCV genotype 1 infection, further analysis was performed in order to evaluate whether the level of palmitic acid could be a predictive factor of the efficacy of anti-HCV treatment. Univariate logistic regression analysis was performed for IL-28B polymorphisms (rs8099917 and rs12979860), amino acid substitution in HCV core 70 and 91, mutated nucleotide sequence of ISDR, and palmitic acid level (Supplementary Table 4). Analysis revealed that palmitic acid level and the ISDR mutation could be significant predictive factors for RVR (P = 0.03 and P = 0.02, respectively). For EVR and SVR, only the palmitic acid level was identified as a significant predictive factor (P = 0.01 for both EVR and SVR). Moreover, multivariate logistic regression analysis revealed that the level of palmitic acid was the only significant contributing factor for EVR (odds ratio = 0.988, 95 % confidence interval 0.979–0.996; P = 0.01); however, it was not a significant factor for RVR or SVR (Table 6).
Table 6

Factors associated with virological response to interferon-based therapy in patients with genotype 1 identified by multivariate analysis

Factor

Category

RVR

EVR

OR (95 % CI)

P value

OR (95 % CI)

P value

Palmitic acid (16:0)

 

0.992 (0.982–0.999)

0.07

0.989 (0.98–0.996)

0.01

Nucleotide sequence of ISDRsa

0–1

1

 

1

 

≥2

5.56 (0.95–38.08)

0.06

5.3 (0.86–48.08)

0.09

Only variables that achieved P < 0.1 on multivariate logistic regression with stepwise method are shown. In SVR, there was no significant factor, which achieved P < 0.1. Total variables include palmitic acid, nucleotide sequence of ISDRs

HCV hepatitis C virus, OR odds ratio, CI confidence interval, RVR rapid virological response, EVR early virological response, ISDR interferon sensitivity-determining region, SVR sustained virological response, IL-28B interleukin-28B, PCR polymerase chain reaction

aIn ISDR, 0–1 was defined as having no amino acid substitutions or one substitution, ≥2 was defined as containing two or more amino acid substitutions

Effect of Fatty Acids In Vitro

The effect of palmitic acid on HCV replication was assessed in vitro using transfected cultured cells expressing H77 (Fig. 1a) or JFH1 (Fig. 1b) HCV clones. In both cell lines, levels of HCV RNA were not altered with the addition of palmitic acid alone. However, anti-HCV effects of IFN and RBV were diminished by addition of palmitic acid (P = 0.028 and P = 0.038 for H77 and JFH1, respectively, by Wilcoxon test). Moreover, other saturated fatty acids (such as myristic acid and stearic acid) and unsaturated fatty acids (such as oleic acid) were assessed. Myristic acid, stearic acid, and oleic acid did not affect HCV replication, and did not alter the treatment effect of IFN and RBV (Supplementary Fig. 2A, 2B).
https://static-content.springer.com/image/art%3A10.1007%2Fs11745-012-3716-8/MediaObjects/11745_2012_3716_Fig1_HTML.gif
Fig. 1

Effects of palmitic acid on in-vitro viral replication and activity of interferon-α plus ribavirin. Levels of HCV RNA were measured after adding interferon-α (IFN-α) and ribavirin (RBV) with or without 100 μM palmitic acid. H77 plasmid-based replication (HCV genotype 1a) (a). JFH1 HCV replication system (HCV genotype 2a) (b). *P = 0.028. **P = 0.038. Data are mean ± standard error of the mean (SEM) of six independent experiments. The Wilcoxon test was used to analyze the data. P < 0.05 was considered statistically significant. ns not significant

Discussion

The present study suggests that low serum levels of palmitic acid could be a predictive factor for virological response to IFN-based therapy in both HCV genotype 1 and 2 infections. In addition, it was also suggested that palmitic acid impairs the anti-HCV effects of IFN and RBV in vitro.

Several studies have shown that HCV core protein disrupts fatty acid homeostasis [9, 26]. HCV core protein has been shown to significantly increase the proportion of C18:1 fatty acids (such as oleic and vaccenic acids), but not palmitic acid, in the livers of patients with HCV infection [9, 10]. Irmisch et al. [27] compared fatty acids in serum of patients with untreated chronic HCV infection with those in treated patients and healthy controls. They showed that women who responded to treatment and healthy controls had significantly higher levels of eicosapentaenoic and arachidonic acid than did untreated HCV patients. There was no significant difference in palmitic acid among the three groups [27]. However, the study did not use pretreatment serum of treated HCV patients and was not a comparison between treated HCV patients who responded to anti-HCV therapy and those who did not. The study compared patients who responded to anti-HCV treatment and untreated patients (including patients who would have responded to anti-HCV treatment if they had received it).

Kapadia et al. and Leu et al. [21, 22] found that both an increase in saturated fatty acids, including palmitic acid, and an increase in monounsaturated fatty acids enhanced HCV replication, whereas increases in polyunsaturated fatty acids such as arachidonic acid suppressed HCV replication in vitro. Huang et al. [28] showed that arachidonic acid inhibited HCV replication by increasing lipid peroxidation, resulting in a decrease in the amount of HCV RNA. In addition, Leu et al. [22] showed that when arachidonic acid was added to IFN-α, a strong synergistic anti-HCV effect was observed in vitro; however, the mechanism of this effect is not well understood.

In the present study, it was found that the concentration of arachidonic acid in the serum of patients infected with HCV was not associated with a change in the effect of IFN-based therapy (data not shown). In fact, the quantity of HCV RNA did not correlate significantly with concentrations of any fatty acids, including palmitic acid, in the serum of HCV patients (data not shown). The present in-vitro data obtained with HCV cell culture systems also indicated that there is no association between saturated fatty acids (including palmitic acid and unsaturated acids such as oleic acid) and HCV replication. However, in this in-vitro system, palmitic acid had an inhibitory effect on IFN-based therapy against HCV. These clinical and in vivo results indicate that inhibitory effects of palmitic acid against IFN-based therapy may be direct effects on hepatocytes.

Another potential explanation for the decreased effect of IFN-based therapy associated with palmitic acid would be the inability of patients to develop effective anti-HCV immunity [29]. It was previously been reported that palmitic acid can induce inflammation and impair the antigen-specific function of dendritic cells (DC) in humans and mice [7]. These conditions might be comparable to chronic HCV infection. Function of DC is impaired in HCV infection [30, 31]. High levels of palmitic acid might further impair the function of DC, reducing anti-HCV immunity. Further studies are needed to identify the mechanisms underlying the effects of palmitic acid, including immunomodulatory effects.

Multiple factors have been reported to be associated with a poor response to IFN-based treatment. Viral factors [including HCV genotype, quantity of HCV RNA, nucleotide sequence of ISDR, and amino acid substitutions in the core region (core aa70 and aa91)] have been shown to affect response to IFN-based therapy [17, 23, 24, 3234]. On the host side, older age, male gender, obesity, insulin resistance or metabolic syndrome, low density lipoprotein, race, and either steatosis or advanced fibrosis on liver biopsy have all been reported as factors associated with poor response to IFN-based therapy [2426, 33, 34]. Recently, IL-28B polymorphism has received attention as a potential factor affecting response to therapy [14].

Based on the present results, previously reported factors that have been shown to influence the effect of IFN therapy were not associated with the level of palmitic acid. Based on the results of univariate and multivariate analyses, the level of palmitic acid was a significant independent predictive factor of response to IFN-based therapy in patients with HCV genotypes 1 and 2. However, a validation study with a larger number of patients with HCV is needed.

In conclusion, this study is the first to report that the serum level of palmitic acid could be a pretreatment predictive factor of virological response with IFN-based therapy in patients with HCV infection. According to the present findings, pretreatment serum concentration of palmitic acid could be used to select patients more likely to respond to IFN-based therapy. It is possible that a special diet or drugs that lower levels of palmitic acid might improve response to IFN-based therapy in patients with HCV infection.

Acknowledgments

This work was supported in part by a Grant-in-Aid for Scientific Research [JSPS KAKENHI 23700907 to T.M.; JSPS KAKENHI 21590848 to Y.H.] from the Japanese Ministry of Education, Culture, Sports, Science and Technology, and a Grant-in-Aid for Scientific Research and Development from the Japanese Ministry of Health, Labor and Welfare to Y.H. We thank Ms. Satomi Yamanaka, Ms. Chie Matsugi and Ms. Sakiko Inoh for excellent technical assistance. The sequence of the probe and primers for the TaqMan assay for detecting rs8099917 was kindly provided by Dr. Yasuhito Tanaka (Department of Virology and Liver Unit, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan), and rs12979860 was provided by Dr. David B. Goldstein (Center for Human Genome Variation, Duke University, Durham, NC, USA). The HCV replication system with pJFH1-full was kindly provided by Dr. Takaji Wakita (Department of Virology II, National Institute of Infectious Diseases, Tokyo, Japan), and that with pT7-flHCV-Rz was provided by Dr. Raymond T. Chung (Gastrointestinal Unit, Massachusetts General Hospital, Boston, MA, USA). We also thank Dr. Francis V. Chisari (Department of Immunology and Microbial Science, The Scripps Research Institute, La Jolla, CA, USA) for providing the Huh7.5.1 human cancer cell lines.

Conflict of interest

None.

Supplementary material

11745_2012_3716_MOESM1_ESM.doc (186 kb)
Supplementary material 1 (DOC 185 kb)
11745_2012_3716_MOESM2_ESM.pdf (291 kb)
Supplementary material 2 (PDF 290 kb)
11745_2012_3716_MOESM3_ESM.pdf (351 kb)
Supplementary material 3 (PDF 351 kb)

Copyright information

© AOCS 2012