Hepatology International

, Volume 7, Issue 2, pp 508–515

Differences in the factors associated with serum viral load between genotypes 1 and 2 in patients with chronic hepatitis C

Authors

    • Department of Gastroenterology and HepatologyJuntendo University Shizuoka Hospital
  • Takuya Genda
    • Department of Gastroenterology and HepatologyJuntendo University Shizuoka Hospital
  • Katsuharu Hirano
    • Department of Gastroenterology and HepatologyJuntendo University Shizuoka Hospital
  • Hironori Tsuzura
    • Department of Gastroenterology and HepatologyJuntendo University Shizuoka Hospital
  • Yoshio Kanemitsu
    • Department of Gastroenterology and HepatologyJuntendo University Shizuoka Hospital
  • Yutaka Narita
    • Department of Gastroenterology and HepatologyJuntendo University Shizuoka Hospital
  • Tetsu Kikuchi
    • Department of Gastroenterology and HepatologyJuntendo University Shizuoka Hospital
  • Katsuyori Ijima
    • Department of Gastroenterology and HepatologyJuntendo University Shizuoka Hospital
  • Takafumi Ichida
    • Department of Gastroenterology and HepatologyJuntendo University Shizuoka Hospital
Original Article

DOI: 10.1007/s12072-011-9273-z

Cite this article as:
Sato, S., Genda, T., Hirano, K. et al. Hepatol Int (2013) 7: 508. doi:10.1007/s12072-011-9273-z
  • 93 Views

Abstract

Purpose

The serum hepatitis C virus (HCV) load is persistently stable in patients with untreated chronic hepatitis C, but its differences between individuals vary widely (above 4 logU/mL). Because serum viral load is an important factor for predicting clinical outcome of interferon-based antiviral therapy, this study was performed to clarify the factors associated with serum viral load in chronic hepatitis C patients.

Methods

We retrospectively analyzed data from 669 chronic hepatitis C patients with HCV genotype 1 or 2 infection. Stepwise regression analysis was used to estimate the relationship between demographic, viral, or biochemical variables and serum viral load.

Results

In univariate analysis, serum lipid profiles, such as total cholesterol, low-density lipoprotein (LDL) and triglyceride levels, and hemoglobin A1c (HbA1c) were correlated with the serum HCV viral load. In multivariate analysis, HCV genotype 1 infection and higher total cholesterol levels were associated with higher viral load. After stratification by HCV genotype, the serum viral load was associated with triglyceride and HbA1c in genotype 1 and with platelet counts and LDL in genotype 2. Histological data (413 patients) showed correlation between severe liver fibrosis and decreased serum viral load in patients with HCV genotype 2 but not genotype 1 infection.

Conclusions

These results suggest that viral kinetics is affected by different host factors for genotypes 1 and 2.

Keywords

GenotypeHepatitis C virusLow-density lipoproteinTriglycerideViral load

Introduction

Infection with the hepatitis C virus (HCV) is a major cause of chronic liver disease. Chronic hepatitis C leads to cirrhosis, liver failure, and hepatocellular carcinoma (HCC). Almost 170 million people worldwide are infected with HCV [1]. In Japan, about 2 million people are infected with HCV, and more than 30,000 patients die from HCC every year [2]. The survival rate and mortality associated with HCV infection are attributable to the progression toward liver failure or HCC.

The current standard therapy for chronic hepatitis C is a combination of pegylated interferon-alpha and the antiviral drug ribavirin. However, a sustained virologic response (SVR) is still accomplished only in about 50% of patients [35]. Previous studies have reported many factors affecting the response to antiviral therapy, including race, age, genetic variation in IL28B, insulin resistance, stage of hepatic fibrosis, adherence both to pegylated interferon and ribavirin, and several viral factors. The viral factors include HCV genotype, viral kinetics during treatment, and amino acid pattern in the interferon sensitivity-determining region of the non-structural protein 5A and in the core protein [68]. Particularly, the serum viral load is one of the most substantial factors that predict clinical outcome. For instance, in patients with genotype 1 and a high viral load (>2 × 106 copies/mL) who were treated with 48 weeks of pegylated interferon and ribavirin, the SVR rate was 46%, but among those with genotype 1 and a low viral load (<2 × 106 copies/mL) who were treated with the same regimen, the rate was 61% [9]. In patients with genotype 2 and a high viral load (>100 kIU/mL) who were treated with 24 weeks of pegylated interferon, the SVR rate was 67%, while those with genotype 2 and a low viral load (<100 kIU/mL) had an SVR of 89% [10].

As previously described, serum viral load is persistently stable over prolonged periods in untreated patients with chronic hepatitis C [11, 12]. However, differences in serum viral load between individuals vary widely (above 4 logU/mL) even among patients with identical genotypes. These observations suggest that interpersonal differences in the serum viral load might be affected by environmental predispositions, such as host demographic and biochemical characteristics. One report indicated that patient characteristics, such as age, gender, body mass index (BMI), and fibrosis score were associated with viral load in HCV genotype 1 [13]. To further understand the regulation of serum viral load, we investigated factors associated with serum HCV viral load in patients infected with HCV genotype 1 or 2, two prevalent genotypes in Japan.

Materials and methods

Study population

This was a retrospective study analyzing 669 adults with chronic hepatitis C admitted to the Department of Gastroenterology and Hepatology at Juntendo University Shizuoka Hospital, Shizuoka, Japan, between April 2005 and December 2009. Eligible subjects had detectable serum HCV-RNA. Exclusion criteria were: (1) positivity for the hepatitis B surface antigen; (2) history or serologic evidence of any other chronic liver diseases, including autoimmune hepatitis, primary biliary cirrhosis, hemochromatosis, and Wilson’s disease; (3) evidence of HCC by ultrasonography or computed tomography; (4) antiviral treatment or immunosuppressive therapy at enrollment; and (5) previous liver transplantation.

This study was approved by the Ethical Committee of Juntendo University Shizuoka Hospital in accordance with the Helsinki Declaration, and written informed consent was obtained from all of the patients.

Laboratory examinations

All laboratory tests were performed for each patient at enrollment. Blood cell counts and serum aspartate transaminase (AST), alanine transaminase (ALT), gamma-glutamyl transpeptidase (GGT), blood glucose, hemoglobin A1c (HbA1c), triglyceride, total cholesterol, high-density lipoprotein (HDL), alpha-fetoprotein (AFP), and des-gamma-carboxy prothrombin (DCP) levels were measured using commercially available assays. Low-density lipoprotein (LDL) was calculated using the Friedewald formula: [LDL (mg/dL)] = [total cholesterol] − [HDL] − ([triglyceride] ÷ 5). In cases where plasma triglyceride exceeds 400 mg/dL, LDL is measured directly. Insulin resistance was determined using homeostasis model assessment (HOMA-IR) method using following equitation: HOMA-IR = fasting insulin (μIU/mL) × fasting glucose (mg/dl) × 0.05551 ÷ 22.5.

Virologic tests

HCV genotype was determined by polymerase chain reaction with the HCV Genotype Primer Kit (Institute of Immunology Co., Ltd., Tokyo, Japan) and classified as genotype 1, genotype 2, or other according to Simmonds’ classification system [14]. Serum HCV viral load was determined by quantitative reverse transcription polymerase chain reaction with COBAS TaqMan HCV Test (Roche Diagnostics, Branchburg, NJ).

Histological evaluation

A total of 413 of the 669 patients (62%) underwent echo-guided percutaneous liver biopsy. Liver biopsy specimens were embedded in paraffin and stained with hematoxylin–eosin, Azan–Mallory, and reticulin silver impregnation to stage the degree of fibrosis. The specimens were evaluated by an experienced pathologist who had no knowledge of the clinical data of the patients. Histological evaluation was based on the METAVIR criteria reported previously [15]. Hepatic fibrosis was defined as follows: F0, no fibrosis; F1, periportal fibrous expansion; F2, portal fibrous widening with bridging fibrosis; F3, bridging fibrosis with lobular distortion; and F4, liver cirrhosis.

Statistical analysis

All analyses were conducted using PASW Statistics software version 18 (SPSS, Chicago, IL, USA). Differences in the clinical backgrounds of patients with HCV genotype 1 or 2 were expressed as median (range) and compared by the Chi-square test for categorical variables or the Mann–Whitney U test for numerical variables. In the univariate analysis, the correlations with viral load were evaluated by Pearson product moment correlation coefficient. In the multivariate analysis, stepwise linear regression analysis was used to identify factors that influence the baseline viral load. P > 0.20 was the significance level for removal from the model and P < 0.1 was the significance level for addition to the model. Body mass index (BMI) was excluded from the multivariate analysis because of the absence of those data in more than 20% of the subjects. Associations between viral load and histological stage were analyzed by the Mann–Whitney U test. P values below 0.05 were considered to be statistically significant.

Results

Characteristics of patients

Table 1 summarizes demographic and biochemical data for all of the 669 patients with chronic hepatitis C, which included 349 male (52%) and 320 female (48%) patients. The median age was 61 years (range 20–87 years). Genotype distribution was as follows: genotype 1, n = 414 (62%); and genotype 2, n = 255 (38%). The serum viral load among all patients was between 1.4 and 7.8 logU/mL. Patients infected with HCV genotype 1 were characterized by older age (P < 0.001), higher viral load (P < 0.001), lower platelet count (P < 0.001), higher AFP level (P = 0.003), and lower DCP level (P = 0.012).
Table 1

Characteristics of all patients according to HCV genotype

 

All patients (N = 669)

Genotype 1 (N = 414)

Genotype 2 (N = 255)

P value

Age (years)

61 (20–87)

63 (20–87)

58 (20–86)

<0.001

Gender (M/F)

349/320

206/208

143/112

0.130

BMI (kg/m2)a

23.2 (13.1–37.3)

23.5 (13.1–36.8)

23.6 (15.3–37.3)

0.924

Viral load (logU/mL)

6.3 (1.4–7.8)

6.4 (3.0–7.8)

6.1 (1.6–7.3)

<0.001

Blood glucose (mg/dL)

102 (56–604)

102 (56–533)

100 (68–604)

0.975

HbA1c (%)

5.0 (3.1–13.5)

5.0 (3.5–10.5)

5.0 (3.1–13.5)

0.780

HOMA-IR

2.40 (0.21–8.90)

2.32 (0.21–8.89)

2.50 (0.42–8.90)

0.616

Triglycerides (mg/dL)

92 (27–1676)

93 (27–516)

89 (33–676)

0.896

Total cholesterol (mg/dL)

167 (62–333)

165 (62–273)

168 (63–333)

0.387

HDL (mg/dL)

55 (12–123)

55 (16–123)

56 (12–105)

0.591

LDL (mg/dL)

88.5 (24–191)

86.0 (25–182)

91.0 (24–191)

0.535

AST (IU/L)

45 (8–4126)

46 (16–4126)

43 (8–682)

0.075

ALT (IU/L)

51 (9–3838)

51 (9–3838)

53 (10–1071)

0.982

GGT (IU/L)

38 (4–1078)

38 (9–749)

37 (4–1078)

0.823

Platelet counts (× 104/μL)

16.7 (1.0–85.9)

16.0 (1.0–85.9)

18.2 (2.9–46.1)

<0.001

AFP (ng/mL)

5 (1–576)

6 (1–307)

5 (1–576)

0.003

DCP (mAU/mL)

19 (7–246)

19 (7–100)

20 (7–246)

0.012

Data are shown as median (range)

P values are for comparisons between genotypes 1 and 2

BMI body mass index, HbA1c hemoglobin A1c, HOMA-IR homeostasis model assessment for insulin resistance, HDL high-density lipoprotein, LDL low-density lipoprotein, AST aspartate transaminase, ALT alanine transaminase, GGT gamma glutamyl transpeptidase, AFP alpha fetoprotein, DCP des-gamma-carboxy prothrombin

Mann–Whitney U test

Chi-square test

aData not available for all patients

Factors associated with the serum HCV viral load

Data from the 669 patients were analyzed by stepwise regression analysis to examine the correlation between the serum viral load and demographic and biochemical variables (such as gender, BMI, age, blood glucose, HbA1c, HOMA-IR, triglycerides, total cholesterol, LDL, HDL, AST, ALT, GGT, platelet counts, AFP, and DCP). Univariate analysis revealed that the levels of total cholesterol (P < 0.001), LDL (P < 0.001), triglycerides (P = 0.018), and HbA1c (P = 0.030) were significantly associated with serum viral load (Table 2). In multivariate analysis, only HCV genotype (P < 0.001) and total cholesterol (P = 0.038) were significantly associated with serum viral load (Table 3). The serum viral load in patients with HCV genotype 1 was significantly higher than that in patients with HCV genotype 2.
Table 2

Univariate analysis of factors correlated with serum viral load

N = 669

Coefficients

P value

Age (years)

0.047

0.226

BMI (kg/m2)a

0.072

0.088

Blood glucose (mg/dL)

0.011

0.769

HbA1c (%)

0.088

0.030

Triglycerides (mg/dL)

0.092

0.018

HOMA-IR

−0.005

0.911

Total cholesterol (mg/dL)

0.162

<0.001

LDL (mg/dL)

0.141

<0.001

HDL (mg/dL)

−0.002

0.960

AST (IU/L)

<0.001

0.993

ALT (IU/L)

−0.005

0.890

GGT (IU/L)

−0.033

0.395

Platelet counts (×104/μL)

0.039

0.312

AFP (ng/mL)

0.020

0.607

DCP (mAU/mL)

−0.001

0.988

For abbreviations, refer to the footnotes of Table 1

Pearson product moment correlation coefficient was used in the univariate analysis

aData not available for all patients

Table 3

Multivariate analysis identifying factors associated with serum viral load

 

Standardized partial regression coefficient

P value

Genotype (2 vs. 1)

−0.315

<0.001

Total cholesterol (mg/dL)

0.096

0.038

HbA1c (%)

0.080

0.066

Triglycerides (mg/dL)

0.083

0.063

Platelet counts (×104/μL)

0.078

0.085

Constant

 

<0.001

Stepwise linear regression analysis was used in the multivariate analysis

P > 0.20 was the significance level for removal from the model and P < 0.10 was the significance level for addition to the model

Factors associated with the serum HCV viral load according to HCV genotype

Because significant differences in the serum viral load were observed between genotypes 1 and 2, we next examined the correlation between the viral load and the variables under stratification by HCV genotype. In patients with genotype 1, univariate analysis showed that serum viral load was significantly correlated with triglycerides (P = 0.018), GGT (P = 0.048), and HbA1c (P = 0.050) (Table 4). In multivariate analysis (Table 5), the serum viral load was independently associated with triglycerides (P = 0.008) and HbA1c (P = 0.032) (Table 5). The regression of viral load versus serum triglyceride levels or HbA1c is shown in Fig. 1.
Table 4

Univariate analysis of factors correlated with serum viral load according to HCV genotype

 

Genotype 1 (N = 414)

Genotype 2 (N = 255)

Coefficients

P value

Coefficients

P value

Age (years)

0.070

0.156

−0.086

0.169

BMI (kg/m2)a

0.019

0.722

0.117

0.096

Blood glucose (mg/dL)

0.070

0.156

−0.038

0.544

HbA1c (%)

0.101

0.050

0.057

0.384

HOMA-IR

−0.017

0.775

0.037

0.633

Triglycerides (mg/dL)

0.119

0.018

0.114

0.073

Total cholesterol (mg/dL)

0.087

0.078

0.275

<0.001

LDL (mg/dL)

0.042

0.432

0.291

<0.001

HDL (mg/dL)

0.001

0.986

0.007

0.923

AST (IU/L)

0.013

0.799

−0.049

0.438

ALT (IU/L)

0.010

0.838

−0.004

0.950

GGT (IU/L)

0.098

0.048

−0.124

0.048

Platelet counts (×104/μL)

−0.124

0.124

0.223

<0.001

AFP (ng/mL)

0.039

0.441

0.008

0.904

DCP (mAU/mL)

0.137

0.358

−0.037

0.576

Pearson product moment correlation coefficient was used to examine the relationship between the baseline viral load and clinical parameters

aData not available for all patients

Table 5

Multivariate analysis identifying factors associated with serum viral load according to HCV genotype

 

Genotype 1

 

Genotype 2

Coefficients

P value

Coefficients

P value

Triglycerides (mg/dL)

0.154

0.008

Platelet count (×104/μL)

0.266

<0.001

HbA1c (%)

0.123

0.032

LDL (mg/dL)

0.163

0.029

Constant

 

<0.001

Constant

 

<0.001

P > 0.20 was the significance level for removal from the model and P < 0.10 was the significance level for addition to the model

Stepwise linear regression analysis was used in the multivariate analysis

https://static-content.springer.com/image/art%3A10.1007%2Fs12072-011-9273-z/MediaObjects/12072_2011_9273_Fig1_HTML.gif
Fig. 1

Regression of serum viral load versus triglyceride (aR2 = 0.014, P = 0.018), HbA1c (bR2 = 0.010, P = 0.050) in HCV genotype 1

In patients with genotype 2, univariate analysis showed that a higher HCV viral load was significantly correlated with increased total cholesterol (P < 0.001), LDL (P < 0.001), and lower platelet counts (P < 0.001) (Table 4). In multivariate analysis, the serum viral load was independently associated with platelet count (P < 0.001) and LDL (P = 0.029) (Table 5). The regression of viral load versus platelet counts or LDL is shown in Fig. 2.
https://static-content.springer.com/image/art%3A10.1007%2Fs12072-011-9273-z/MediaObjects/12072_2011_9273_Fig2_HTML.gif
Fig. 2

Regression of serum viral load versus platelet counts (aR2 = 0.054, P < 0.001), LDL (bR2 = 0.085, P < 0.001) in HCV genotype 2

Serum HCV viral load and hepatic fibrosis stage

Because it is largely known that platelet counts reflect the extent of hepatic fibrosis [16], we investigated the relationship between viral load and hepatic fibrosis stage according to HCV genotype (Fig. 3). Histological data were available for 413 of the patients (62%). A total of 215 of the evaluated patients (52%) had mild hepatic fibrosis (F0–F1), and 198 (48%) had severe hepatic fibrosis (F2–F4). In genotype 1 infection, 128 patients (51%) had mild hepatic fibrosis (F0–F1), and 122 patients (49%) had severe hepatic fibrosis (F2–F4). Their median HCV viral loads were 6.4 logU/mL (range 3.7–7.8 logU/mL) and 6.4 logU/mL (range 4.1–7.8 logU/mL), respectively. No significant difference was observed. In genotype 2 infection, 87 patients (53%) had mild hepatic fibrosis (F0–F1), and 76 patients (47%) had severe hepatic fibrosis (F2–F4). Their median HCV viral loads were 6.3 logU/mL (range 1.6–7.3 logU/mL) and 5.9 logU/mL (range 3.5–6.7 log/mL), respectively, showing statistically significant differences. The difference in serum viral load between genotypes 1 and 2 was observed only in patients with severe fibrosis (F2–F4), but not in patients with mild fibrosis (F0–F1).
https://static-content.springer.com/image/art%3A10.1007%2Fs12072-011-9273-z/MediaObjects/12072_2011_9273_Fig3_HTML.gif
Fig. 3

Association between serum viral load and histological stage. *Statistical significance between two groups (P < 0.001)

Discussion

At present, the recommended treatment for chronic hepatitis C is administration of pegylated interferon-alpha in combination with ribavirin. Many factors have recently been reported to predict the outcome of treatment, such as HCV genotype, treatment duration, patients’ gender and age, etc. Among them, serum viral load has been revealed as one of the critical predictors, despite of HCV genotype [9, 10]. In general, serum HCV viral load is persistently stable, and its intrapersonal differences are small in patients with chronic hepatitis C [11, 12], but its interpersonal differences are wide-ranging even among patients infected with identical genotypes. The interpersonal difference in serum viral load was above 6 logU/mL in our cohort. Because this observation suggests the involvement of host characteristics in serum HCV viral load, we analyzed a large number of patients with chronic hepatitis C to clarify their association.

Serum HCV viral load is thought to be determined by viral kinetics, such as viral replication, secretion from hepatocytes, and clearance from blood flow. Evidence involving the molecular mechanisms of HCV viral kinetics has been accumulated recently. HCV replication requires the product of the host mevalonate pathway which synthesizes cholesterol from acetyl-CoA in hepatocytes [17, 18]. The viral replication complex is formed on the lipid raft, a cholesterol-rich membrane structure [19, 20]. Infectious HCV particles bind to host beta-lipoproteins, such as LDL and VLDL during HCV assembly and release from infected hepatocytes [21, 22]. The LDL receptor and the human scavenger receptor class B type 1 receptor, HDL receptor, play an important role in virus cell entry into hepatocytes [2325]. Thus, reliance on the host lipid metabolic pathway for viral kinetics seems to be a unique feature of HCV, and is suggestive of the relationship between serum lipid profiles and viral kinetics in patients with chronic hepatitis C. Indeed, our present analysis of 669 patients with chronic hepatitis C identified a significant correlation between serum HCV viral load and serum total cholesterol. Until now, several studies have reported an association between HCV infection and alterations in host lipid profiles [2628], but a correlation between host lipid profiles and serum HCV viral load was shown in only a few reports [29, 30].

In addition to an association between viral load and total cholesterol, we identified a close relationship between serum viral load and HCV genotype. In patients with HCV genotype 1, a significantly higher serum HCV viral load was shown in comparison to that with genotype 2. Although consistent results were obtained in several previous studies [3133], the reason why the viral load depends on HCV genotype remained largely unknown. Therefore, we analyzed the correlation between patient characteristics and serum viral load under stratification by HCV genotype. In patients with genotype 1, triglycerides and HbA1c were identified as factors associated with serum viral load. On the other hand, LDL and platelet count were identified in patients with genotype 2. Thus, different lipid profiles and biochemical factors were associated with serum viral load between genotypes 1 and 2. Interestingly, in addition to the difference in the associated factors, the R2 values of regression analysis were larger in patients with genotype 2 than in those with genotype 1, suggesting that the viral load in genotype 2 was more sensitive to the change in host characteristics than that in genotype 1. These differences might be involved in the differences between serum viral loads in genotypes 1 and 2.

It remains largely unknown why different factors were associated with serum viral load between genotypes 1 and 2. Many studies have reported the possible involvement of the host lipid metabolic pathways in HCV viral kinetics. Among them, VLDL, in which triglycerides are mainly presented, is known to play a unique role in the HCV life cycle. Several studies have shown that HCV particles were attached to or incorporated into VLDL, and that HCV production by hepatocytes depended on the assembly and secretion of VLDL [34]. Our clinical observation that serum triglycerides level was correlated with genotype 1, HCV viral load is consistent with these results. On the other hand, a recent report stated that serum LDL level was associated with IL28B genetic polymorphisms [35]. IL28B polymorphism is not only a predictor of treatment responses but also associated with pretreatment HCV viral load, suggesting that IL28B polymorphism has a role in the regulation of host baseline interferon-stimulated gene expression [36]. Taken together, a higher serum LDL level possibly indicates a lower endogenous interferon response to HCV. It is known that patients with genotype 2 infection are much more sensitive to interferon-based antiviral therapy than those with genotype 1 infection. The correlation between genotype 2 HCV viral load and serum LDL levels might be explained by its high sensitivity to endogenous interferon response.

Analysis of patients with histology-proven genotype 2 revealed that the serum HCV viral load was lower in patients with severe hepatic fibrosis than in those with mild hepatic fibrosis. Because it is widely accepted that platelet counts are inversely correlated with the degree of hepatic fibrosis, this result was consistent with our observation that the serum viral load was associated with the platelet count in patients with genotype 2. The relationship between the HCV viral load and the severity of liver disease or progression to HCC has remained controversial. Several previous studies revealed that the HCV viral load declines with the deterioration of liver disease, particularly in the development of HCC [32, 37]. Other studies reported that the viral loads did not correlate with stage of liver disease [31, 38], but these studies were conducted on small cohorts of patients or on the patients mainly infected with genotype 1b. According to our results, the serum HCV viral load is correlated with the degree of liver fibrosis at least in patients infected with genotype 2. Importantly, the difference in viral load between genotypes 1 and 2 was observed only in patients with severe fibrosis, not in those with mild fibrosis. A decrease in genotype 2 viral load in association with progression of liver fibrosis could explain the difference in viral loads between genotypes 1 and 2.

In conclusion, the serum viral load was associated with the host lipid profile and HCV genotype, at least in patients with genotype 1 or 2. In addition, different associations of lipid profiles and liver fibrosis with serum viral load were demonstrated between genotypes 1 and 2. These differences might be involved not only in a lower viral load but also higher susceptibility to treatment in patients with genotype 2 than in those with genotype 1. Further analysis of these differences will provide us with more effective treatment strategies according to HCV genotype.

Copyright information

© Asian Pacific Association for the Study of the Liver 2011