Insulin resistance is associated with steatosis in nondiabetic patients with genotype 1 chronic hepatitis C

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

Conflicting data exist regarding the relationship between hepatitis C virus genotype 1 and hepatic steatosis as well as the latter's role in the progression of fibrosis and treatment response. We assessed factors associated with hepatic steatosis in genotype 1 chronic hepatitis C and the impact of hepatic fat on fibrosis development and interferon responsiveness. Two hundred ninety-one non-diabetic patients with genotype 1 chronic hepatitis C were examined for the presence of steatosis and its correlation with clinical, virological, and biochemical data, including insulin resistance (IR), evaluated by the homeostasis model assessment (HOMA) score. Steatosis was graded as mild (1%-20% of hepatocytes involved), moderate (21%-40% of hepatocytes involved), and severe (>40% of hepatocytes involved). Steatosis was mild in 110 of 291 (37.8%) and moderate/severe in 55 of 291 (18.9%) subjects. By logistic regression, moderate/severe steatosis was independently associated with the female sex (odds ratio [OR] 2.74; 95% CI 1.40-5.35), high γ-glutamyltransferase levels (OR 1.52; 95% CI 1.22-1.91), and HOMA-score (OR 1.076; 95% CI 1.001-1.26). By logistic regression, moderate/severe steatosis (OR 2.78; 95% CI 1.21-6.4), and platelet counts (OR 0.97; 95% CI 0.96-0.98) were independent predictors of advanced fibrosis. Patients with moderate/severe steatosis had an OR of 0.52 (95% CI 0.30-0.90) for sustained virological response compared with patients with mild/absent steatosis. In conclusion, in nondiabetic European patients with genotype 1 hepatitis C at low risk for the metabolic syndrome, the prevalence of steatosis was nearly 60%. IR is a risk factor for moderate/severe steatosis, especially in men. Moderate/severe steatosis has clinical relevance, being associated with advanced fibrosis and hyporesponsiveness to antiviral therapy. (HEPATOLOGY 2006;43:64–71.)

Hepatic steatosis is a common histological finding in chronic hepatitis C virus (HCV) infection, occurring in 30% to 70% of such patients. 1, 2 The biological mechanism underlying steatosis in HCV infection is not definitively understood and is considered to be multifactorial. Recent evidence suggests that body mass index (BMI), HCV genotype 3, and fibrosis are independently associated with steatosis on multivariate analysis.2–5 Clinical data from patients with genotype 3 infection have shown that steatosis resolves in patients who achieve sustained virological response (SVR) after therapy, providing direct evidence of a steatogenic effect of the virus.6, 7 One of the mechanisms proposed to explain this effect is the capability of the HCV genotype 3 core protein to inhibit secretion of very-low-density lipoprotein in the liver and consequently to induce steatosis together with hypobeta-lipoproteinemia.8, 9

Data on steatosis in patients with genotype 1 infection are less clear because different metabolic factors such as hypertension, obesity, and alterations in hepatic lipid and carbohydrate metabolism 2, 5, 10 hampered the assessment of a relationship between steatosis and HCV genotype 1. In patients infected with HCV genotype 1, viral eradication has no effect on steatosis,6 and no modifications of the secretion of very-low-density lipoprotein in the liver have been reported. Nevertheless, studies with constructs containing HCV core encoding sequences from genotype 1 isolates11–14 have suggested a direct effect of the HCV genotype 1 core protein in inducing steatosis. Recent clinical15 and experimental16 data showed a correlation between HCV genotype 1 and the development of insulin resistance (IR), supporting the earlier findings of the association between type 2 diabetes mellitus and chronic hepatitis C.

Whatever the mechanisms responsible for hepatic steatosis in HCV genotype 1 infection, the crucial clinical question about steatosis in hepatitis C is whether it may enhance the progression of fibrosis. Mounting experimental 17 and clinical evidence2–5, 18, 19 has shown that hepatic steatosis accelerates the development and progression of fibrosis in chronic hepatitis C.

Another unsolved issue concerns the impact of liver steatosis on the response to antiviral treatment. Several pretreatment features have been found useful in identifying patients with chronic hepatitis C who might benefit from peginterferon plus ribavirin treatment. These included sex, age, weight, viral load, HCV genotype 1, and degree of fibrosis. 20–24 The degree of steatosis appears to reduce the likelihood of achieving SVR in genotype 1 infected patients,7, 20, 24 although current data are limited. Finally, a recent paper by Romero-Gómez et al.25 found that IR is an independent predictor of SVR in patients with chronic hepatitis C who are treated with peginterferon and ribavirin.

We analyzed data on a large cohort of European nondiabetic patients with genotype 1 chronic hepatitis C from a randomized controlled trial (RCT) of peginterferon and ribavirin therapy, 20 in an attempt to assess the role of IR in hepatic steatosis. In this study, we have also examined the correlation between steatosis and fibrosis and the impact of hepatic fat on treatment response.

Abbreviations

HCV, hepatitis C virus; BMI, body mass index; SVR, sustained virological response; IR, insulin resistance; RCT, randomized controlled trialALT: alanine aminotransferase; WBC, white blood count; GGT, γ-glutamyltransferase;; HOMA: homeostasis model assessment; PCR, polymerase chain reaction.

Patients and Methods

Patients.

The current study was designed to analyze data from 311 consecutive patients with genotype 1 chronic HCV recruited from eight tertiary referral liver units and enrolled in an Italian multicenter RCT of peginterferon alfa-2b plus ribavirin. 20

Inclusion criteria were as follows: (1) previously untreated HCV RNA–positive patients between the ages of 18 and 65 years, with alanine aminotransferase (ALT) values 1.5 above the upper normal limit; (2) liver biopsy performed within 6 months before enrollment and a diagnosis of chronic hepatitis with any degree of fibrosis; (3) hemoglobin ≥ 13 g/dL for men, ≥ 12 g/dL for women, white blood cell (WBC) count > 3.000/mm3, granulocyte count > 1.500/mm3, platelet count 80.000/mm3, bilirubin, albumin, and serum creatinine levels within normal limits; (4) self-reported complete abstinence from alcohol during antiviral treatment and posttreatment follow-up. A questionnaire regarding past and current alcohol intake and drug use or abuse was administered before treatment. Daily ethanol consumption was categorized as follows: (1) < 20 g, (2) 20-39 g, (3) 40-59 g, (4) 60-79 g, or (5) ≥80 g/d. All patients were asked to completely abstain from alcohol during antiviral treatment and posttreatment follow-up. During follow-up visits, maintenance of abstinence was verified by questioning.

Patients were excluded if they had: (1) advanced cirrhosis, that is, large esophageal varices (F2 or F3), history of gastrointestinal bleeding, ascites, or encephalopathy; (2) hepatocellular carcinoma; (3) fasting glucose level > 6.2 mmol/L or antidiabetic treatment; (4) anti-HIV or hepatitis B surface antigen positivity; (5) parenteral drug addiction if not abstaining for at least 2 years; (6) any other contraindications to interferon or ribavirin.

The current study was performed in accordance with the principles of Good Clinical Practice, the principles of the Declaration of Helsinki, and its appendices, and local and national laws.

Clinical and Laboratory Assessment.

The following data were collected at the time of liver biopsy: age, sex, weight, height. BMI was calculated as weight in kilograms/height in square meters. Patients with a BMI of 18.5 to 24.9 were classified as normal, those with a BMI of 25 to 29.9 as overweight, those with a BMI of 30 to 34.9 as moderately obese, and those with a BMI ≥ 35 as severely obese.

An overnight (12 hours) fasting blood sample was drawn at the time of liver biopsy to determine the serum levels of ALT, γ-glutamyltransferase (GGT), alkaline phosphatase, cholesterol, triglycerides, ferritin, plasma glucose concentration, and platelet count. Serum insulin was determined by a two-site enzyme ELISA (MERCODIA Insulin ELISA, Uppsala, Sweden). The detection limit was less than 1 μU/mL. IR was determined by the homeostasis model assessment (HOMA) method by using the following equation 26: Insulin resistance (HOMA-IR) = Fasting insulin (μU/mL) × fasting glucose (mmol/L)/22.5. HOMA-IR has been validated in comparison with euglycemic/hyperinsulinemic clamp technique in both diabetic and nondiabetic patients.27

Serum HCV RNA was measured at each center by a qualitative polymerase chain reaction (PCR) assay (Cobas Amplicor HCV Test version 2.0; limit of detection: 50 IU/mL) before treatment, during treatment, and 24 weeks after treatment. Quantification of serum HCV RNA levels before treatment was performed at a central laboratory (Policlinico San Matteo, University of Pavia) by Versant HCV RNA 3.0 bDNA (Bayer Co., Tarrytown, NY) and expressed in IU/mL. Genotyping was performed by INNO-LiPA, HCV II (Bayer). SVR was defined as the absence of detectable serum HCV RNA by PCR 24 weeks after cessation of therapy.

Assessment of Histology.

Slides of liver biopsy specimens obtained at study entry were coded and read by a single pathologist (E.M. at Niguarda Hospital, Milano, Italy) who was unaware of patients' identities and treatment regimens. A minimum length of 15 mm liver biopsy specimen or the presence of at least 10 complete portal tracts were required. 28 Liver biopsies were classified according to the numerical scoring system of Ishak.29 Eachspecimen was reviewed under low (2-4×) and higher (10×) magnification. The percentage of hepatocytes containing macrovesicular fat was determined for each 10× field. An average percentage steatosis was then determined for the entire specimen. Steatosis was assessed as the percentage of hepatocytes containing macrovesicular fat droplets. It was graded as follows: 0 = no steatosis; 1 = mild: 1% to 20% of hepatocytes affected; 2 = moderate: 21% to 40% of hepatocytes affected; 3 = severe: >40% of hepatocytes affected. Finally, steatosis was graded according to Brunt's scoring system30 based on the percentage of hepatocytes in the biopsy involved, as follows: 0 (no steatosis), 1 (<33% of hepatocytes affected), 2 (33%-66% of hepatocytes affected), or 3 (>66% of hepatocytes affected).

Statistical Methods.

Continuous variables were summarized as mean ± SD and categorical variables as frequency and percentage. Multiple logistic regression models were used to assess the relationship of steatosis, fibrosis, and SVR with demographic, virological, and metabolic characteristics of the patients. In the first model, the dependent variable was steatosis coded as 0 = absent or mild (0% to 20% of hepatocytes affected) or 1 = moderate or severe (>20% of hepatocytes affected). As candidate risk factors for moderate/severe steatosis, we selected age, sex, weight, BMI, baseline ALT, platelet count levels, GGT levels, ferritin, blood glucose, insulin, cholesterol, triglycerides, HOMA score, pretreatment HCV-RNA levels (<8 × 105 IU/mL vs. ≥8 × 105 IU/mL), fibrosis score (1-3 vs. 4-6), and activity score (1-8 vs. 9-18). Multiple linear regression analysis was performed to identify independent predictors of steatosis as continuous dependent variable.

In the second model, the dependent variable was fibrosis coded as 0 = 1 to 3 points in the fibrosis score; 1 = 4 to 6 points in the fibrosis score. We considered as explanatory variables age, sex, weight, BMI, baseline ALT, platelet count levels, GGT levels, ferritin, blood glucose, insulin, cholesterol, triglycerides, HOMA score, pretreatment HCV RNA levels (<8 × 105 IU/mL vs. ≥8 × 105 IU/mL), steatosis (absent/mild vs. moderate/severe), and activity score (1-8 vs. 9-18).

In the third model, the dependent variable was SVR coded as 0 = non-SVR; 1 = SVR. On the basis of experience gathered from previously untreated patients, we selected as candidate predictors of SVR age, sex, weight, BMI, baseline ALT, platelet count levels, GGT levels, ferritin, blood glucose, insulin, cholesterol, triglycerides, HOMA score, pretreatment HCV-RNA levels (<8 × 105 IU/mL vs. ≥8 × 105 IU/mL), steatosis (absent/mild vs. moderate/severe), fibrosis score (1-3 vs. 4-6), and activity score (1- 8 vs. 9-18).

Variables found to be associated with the dependent variables on univariate logistic regression at a probability threshold < 0.10 were included in multivariate logistic regression models. Regression analysis was performed using PROC LOGISTIC and PROG REG subroutine in SAS (SAS Institute, Inc., Cary, NC). 31

Results

Patients' Features and Histological Findings.

Of 311 patients enrolled in the previously published RCT, 20 (6.4%) were excluded from this analysis because they had an established diagnosis of diabetes mellitus or a fasting glucose level > 6.2 mmol/L.

Baseline characteristics of all the 291 nondiabetic patients included are shown in Table 1, according to grade of steatosis. Overall, mean values for serum glucose, cholesterol, and triglycerides were within the normal range.

Table 1. Demographic, Laboratory, and Histological Features of 291 Nondiabetic Patients With Genotype 1 Chronic Hepatitis C, According to Steatosis Grade
VariableAll n = 291Grade of Steatosis
Absent n = 126Mild n = 110n = 55
  1. Abbreviations: ULN, upper limit of normal; HCV-RNA, hepatitis C virus ribonucleic acid; IU, international units.

Mean age (yrs), mean ± SD49.7 ± 11.045.9 ± 11.252.3 ± 9.952.9 ± 10.3
Age (yrs), n (%)    
 ≤ 50121 (41.5)71 (56.3)32 (29)18 (32.7)
 > 50170 (58.4)55 (43.6)78 (71)37 (67.3)
Sex, n (%)    
 Male180 (61.8)79 (62.6)74 (67.2)27 (49)
 Female111 (38.2)47 (37.3)36 (32.8)28 (51)
Mean weight (kg)70.4 ± 12.468.5 ± 12.573.0 ± 12.469.5 ± 12.1
Weight (kg)    
 ≤ 70154 (52.9)75 (59.5)47 (42.7)32 (58.8)
 > 70137 (47.1)51 (40.4)63 (57.3)23 (41.2)
Mean body mass index (kg/m2)24.8 ± 3.324.0 ± 3.325.5 ± 3.425.1 ± 3.0
Body mass index (kg/m2)    
 < 25163 (56)82 (65)52 (47.2)29 (52.7)
 25–29.9106 (36.4)39 (32.5)46 (41.8)21 (38.1)
 ≥ 3022 (7.5)5 (3.9)12 (10.9)5 (9.0)
Alanine aminotransferase—UNL3.0 ± 2.03.0 ± 2.32.9 ± 1.83.2 ± 1.7
Platelet count × 103/mmc189.3 ± 58.7198.8 ± 59.4181.5 ± 59.2183.3 ± 54.6
γ GGT—UNL1.8 ± 1.31.5 ± 1.21.6 ± 0.92.4 ± 1.9
Ferritin (ng/mL)306.1 ± 207.5256.9 ± 185.2326.3 ± 203.7330.1 ± 195.3
Blood glucose (mmol/L)4.65 ± 0.844.57 ± 0.824.63 ± 0.894.87 ± 0.76
Insulin (μU/mL)11.0 ± 11.410.95 ± 129.48 ± 5.814.8 ± 17.6
HOMA score2.74 ± 2.372.66 ± 2.592.50 ± 1.393.41 ± 3.2
Cholesterol (mg/dL)192.6 ± 37.4196.2 ± 38.9192.9 ± 36.9184.5 ± 34.2
Triglycerides (mg/dL)108.0 ± 46.7103.1 ± 43.9108.5 ± 46.0117.4 ± 52.3
Log10 serum HCV RNA (IU/mL)12.7 ± 2.212.7 ± 2.412.9 ± 1.812.4 ± 2.3
Histology at biopsy    
 Stage of fibrosis    
  1–3232 (79.7)110 (87.3)86 (78.1)36 (65.4)
  4–659 (20.3)16 (12.7)24 (21.8)19 (34.5)
 Grade of inflammation    
 Minimal-mild (1–8)    
 Moderate/severe (9–18)247 (84.8)111 (88)89 (80.9)47 (85.5)
 44 (15.1)15 (11.9)21 (19)8 (14.5)

The mean biopsy length was 19.3 ± 5.1 mm. Twenty percent of patients (59/291) had fibrosis ≥ 4 by Ishak score. The proportion of patients with moderate/severe necroinflammation (Ishak 9-18) was low (15.1%). Overall, 165 of 291 patients (56.7%) had histological evidence of steatosis. Steatosis was grade 1 (mild = 1%-20%) in 110 of 291 (37.8%) patients, grade 2 (moderate = 21%-40%) in 46 of 291 (15.8%) patients, and grade 3 (severe > 40%) in 9 of 291 (3%) patients. According to Brunt's classification, steatosis was grade 1 in 149 (51.2%) patients, grade 2 in 14 (4.8%) patients, and grade 3 in 1 (0.003%) patient. In view of the small number of patients with grade 3 steatosis of both the scoring systems, patients with grades 2 and 3 were combined for statistical analysis.

Factors Associated With Moderate/Severe Hepatic Steatosis.

Univariate and multivariate analyses were performed to identify predictors of moderate/severe steatosis. Univariate comparison of variables between patients with moderate/severe steatosis and those with absent/mild steatosis is reported in Table 2. Female sex, high baseline values of GGT, HOMA score, fibrosis ≥ 4, age, and triglycerides were associated with moderate/severe steatosis (P < .10). Multivariate logistic regression analysis (Table 3) showed that the following were independent and significant risk factors for moderate/severe steatosis: female sex, high GGT levels, and HOMA score. Fibrosis ≥ 4, age, and triglycerides were no longer significant by multivariate analysis. The estimated risk for moderate/severe steatosis in a hypothetical patient with HOMA score at the 75th percentile was 16% higher as compared with a patient with a median HOMA score. A significant interaction was seen between sex and HOMA score (P = .02), which became of marginal statistical significance after adjustment for the other independent variables (P = .052).The model for the independent predictors of steatosis as continuous variable by multiple linear regression analyses included female sex (P = .002), high GGT levels (P = .0001), HOMA score (P = .01), and BMI (P = .01).

Table 2. Univariate Analysis of Risk Factors for Moderate/Severe Liver Steatosis in 291 Nondiabetic Patients With Genotype 1 Chronic Hepatitis
VariableAbsent/mild n = 236Moderate/severe n = 55P
  1. Abbreviations: ULN, upper limit of normal; HCV-RNA, hepatitis C virus ribonucleic acid; IU, international units.

Mean age (y)48.8 ± 11.152.9 ± 10.6.07
Sex  .03
 Male153 (85)27 (15) 
 Female83 (74.7)28 (25.2) 
Weight (kg)70.5 ± 12.669.5 ± 12.1.58
Body mass index (kg/m2)24.6 ± 3.425.1 ± 3.0.28
Alanine aminotransferase—UNL2.9 ± 2.13.2 ± 1.7.41
Platelet count × 103/mmc191.7 ± 60.4183. ± 54.6.79
γ Glutamyl transferase—UNL1.5 ± 1.12.4 ± 1.9.0001
Ferritin (ng/mL)285.2 ± 196.2330.1 ± 195.3.12
Blood glucose (mmol/L)4.6 ± 0.854.87 ± 0.76.06
Insulin (μU/mL)10.2 ± 9.7414.8 ± 17.6.06
HOMA score2.58 ± 2.13.41 ± 3.2.03
Cholesterol (mg/dL)195.2 ± 42.5184.5 ± 34.2.23
Triglycerides (mg/dL)101.3 ± 41117.4 ± 52.3.07
Log 10 serum HCV RNA (IU/mL)12.8 ± 2.112.4 ± 2.3.25
Histology at biopsy   
 Stage of fibrosis  .004
  1–3196 (84.4)36 (15.5) 
  4–640 (67.7)19 (32.2) 
 Grade of inflammation200 (80.9)47 (19).88
  Minimal-Mild (1–8)36 (81.8)8 (18.1) 
  Moderate/Severe (9–18)   
Table 3. Logistic Regression Model to Predict Moderate/Severe Liver Steatosis in 291 Nondiabetic Patients With Genotype 1 Chronic Hepatitis
VariablePOR (95% CI)
  1. Abbreviation: ULN, upper limit of normal.

Sex:  
 0: male.00312.74 (1.40–5.35)
 1: female  
γ-GGT (per ULN).00021.52 (1.22–1.91)
HOMA score (per unit).0461.076 (1.001–1.26)
Stage of fibrosis:.221.140 (0.92–1.42)
 0: 1–3  
 1: 4–6  
Age (y).391.014 (0.98–1.046)
Triglycerides (mg/dL).251.005 (0.99–1.013)

Replacing moderate/severe steatosis with Brunt's scoring system as the dependent variable in the multivariate analysis, we obtained a model including high GGT levels (OR 1.26; 95% CI 1.05-1.52) as the only significant predictor of the grade of steatosis.

Multivariate analysis performed on the subset of of 269 nonobese patients (BMI < 30), again confirmed female sex (OR 3.26; 95% CI 1.52-6.0), high GGT levels (OR 1.56; 95% CI 1.24-1.95), and HOMA score (OR 1.06; 95% CI 1.001-1.27) as the only significant predictors of moderate/severe steatosis. The same analysis performed in the subset of 163 lean patients (BMI < 25) also showed that female sex (OR 4.81; 95% CI 1.80-12.8) and high GGT levels (OR 1.48; 95% CI 1.09-2.0) were significant predictors of moderate/severe steatosis whereas HOMA score showed a similar effect size but not statistical significance (OR 1.05; 95% CI 0.80-1.15).

Because of the close association of sex with the presence and severity of steatosis, we performed separate analyses to evaluate any possible difference between women and men in risk factors for moderate/severe steatosis. Among the 111 women, high GGT levels (OR 1.98; 95% CI 1.18-3.31), and older age (OR 1.055; 95% CI 1.001-1.112) were the only factors significantly associated with moderate/severe steatosis. Of 28 womenwith moderate/severe steatosis, 25 (89%) were postmenopausal women over the age of 55 years.

In the 180 men, multivariate analysis showed the following as independent risk factors for moderate/severe steatosis: high GGT levels (OR 1.39; 95% CI 1.08-1.77), HOMA score (OR 1.22; 95% CI 1.002-1.52), and fibrosis score ≥ 4 (OR 3.16; 95% CI 1.27-7.85). Age was not a significant risk factor for steatosis in men. Tto assess whether IR is associated with steatosis even in the absence of obesity in men, a multivariate analysis was performed on the subset of 165 men with a BMI < 30. This analysis again confirmed high GGT levels (OR 1.35; 95% CI 1.05-1.73), HOMA score (OR 1.30; 95% CI 1.003-1.68), and fibrosis score ≥ 4 (OR 2.75; 95% CI 1.03-7.31) as significant risk factors for moderate/severe steatosis. In men segregated with respect to BMI < 25, the prevalence of moderate/severe steatosis was 77% (7/9) in the group with HOMA score > 3 and 48.7% (38/78) in the group with HOMA score ≤ 3 (P = .09).

Finally, multivariate analysis excluding the 25 men with cirrhosis (i.e., Ishak score 5 or 6) also showed a similar effect-size of the HOMA score but marginal statistical significance: high GGT levels (OR 1.36; 95% CI 1.05-1.77) and HOMA score (OR 1.25; 95% CI 0.99 -1.60).

Impact of Steatosis on Fibrosis and on SVR.

By univariate analysis, older age, high baseline ALT, ferritin, GGT, low platelet counts, high triglycerides, high activity grade, and moderate/severe steatosis were significantly associated with fibrosis ≥ 4 (Table 4). By multivariate analysis, moderate/severe steatosis (OR 2.78; 95% CI 1.21-6.4), and platelet counts (OR 0.97; 95% CI 0.96-0.98) were the only independent predictors of severity of fibrosis (Table 4).

Table 4. Predictors of Severe Fibrosis (Score ≥ 4 by Ishak) in 291 Nondiabetic Patients With Genotype 1 Chronic Hepatitis
PredictorUnivariate AnalysisMultivariate Analysis
OR (95% CI)POR (95% CI)P value
  1. Abbreviations: ULN, upper limit of normal; HCV-RNA, hepatitis C virus ribonucleic acid; IU, international units.

Age (y)1.059 (1.026–1.093).00041.020 (0.98–1.06).32
Sex0.650 (0.35–1.21).17 
Male vs. female    
Weight (kg)1.013 (0.99–1.03).27 
Body mass index (kg/m2)1.050 (0.96–1.14).23 
Alanine aminotransferase—UNL1.140 (1.004–1.29).041.023 (0.87–1.12).78
Platelet count × 103/mmc0.970 (0.96–0.98).00010.970 (0.96–0.98).0001
γ GGT—UNL1.430 (1.18–1.75).00031.150 (0.90–1.47).24
Ferritin (ng/mL)1.002 (1.001–1.004).00071.001 (0.99–1.002).39
HOMA score1.043 (0.93–1.16).44 
Cholesterol (mg/dL)0.990 (0.98–1.001).090.990 (0.98–1.006).39
Triglycerides (mg/dL)1.008 (1.001–1.015).0261.010 (0.99–1.021).43
Serum HCV RNA (IU/mL)    
<8 × 105 vs ≥ 8 × 1051.042 (0.58–1.85).89  
Histology at biopsy    
 Steatosis    
  Absent/mild vs. moderate/severe2.580 (1.34–4.96).0042.500 (1.064–5.88).035
 Grade of inflammation    
  Minimal-mild vs. moderate/severe2.380 (1.18–4.92).0151.140 (0.47–2.74).76

Logistic regression analysis showed that patients with moderate/severe steatosis had an OR of 0.52 (95% CI 0.30-0.90) for SVR compared with patients with absent/mild steatosis. The other factors independently associated with a low likelihood of SVR were: age > 50 years (OR 0.41; 95% CI 0.23-0.73), fibrosis stage ≥ 4 (OR 0.37; 95% CI 0.15-0.88), and high GGT levels (OR 0.71; 95% CI 0.53-0.94). Again, logistic regression analysis that simultaneously took into account the effect of moderate/severe steatosis as well as age, fibrosis stage, GGT levels, and type of treatment (standard vs. pegylated interferon) confirmed that the effect of moderate/severe steatosis on SVR was not affected by this adjustment (OR of 0.54; 95% CI 0.26-0.96).

Discussion

This study of nondiabetic patients who are at low risk for the metabolic syndrome infected by HCV genotype 1 and who were enrolled in an RCT of peginterferon and ribavirin therapy shows that overall 57% of patients had histological evidence of steatosis, mostly of mild degree. This prevalence was not vastly different from that reported in other observations, ranging from 30% to 70%. 1 The exclusion of patients with diabetes and overt alcohol abuse and the low number of both obese patients and patients with associated dyslipidemia suggest that factors other than overweight, alcohol, or alterations in hepatic lipid and carbohydrate metabolism are involved in the steatogenic process in chronic hepatitis C caused by genotype 1 infection. IR, female sex, and elevated pretreatment GGT levels were independent risk factors for moderate/severe steatosis.

The association we found between IR and moderate/severe steatosis supports the concept that IR is a risk factor for fatty liver. In men, HOMA-IR remained a significant risk factor for moderate/severe steatosis even after the exclusion of subjects with a BMI ≥ 30 and those with cirrhosis, both of whom may have been associated with IR. The lack of correlation between IR and steatosis observed in the studies by Hui et al. 15 and by Muzzi et al.32 was probably caused by the confounding effect of BMI in the former15 and by alcohol consumption in the latter.32 We confirmed the observation by Romero-Gomez et al.,3 who failed to demonstrate a significant correlation between viral load and degree of hepatic steatosis in patients infected with genotype 1. These data suggest that steatosis in the liver of these patients is independent from viral replication, whereas it may be linked to virus-induced metabolic abnormalities, such as IR. Further support to this hypothesis is lent by the fact that fatty liver disappears only in a low proportion of patients with genotype 1 chronic hepatitis after SVR.33 This result is at variance with genotype 3 infection, in which steatosis resolves in most patients after eradication of HCV through treatment, unless metabolic factors are present.33

An important novel finding of this work is that women develop more moderate/severe steatosis than men and that the pathway to steatosis may be different in men and in women. The sex-associated differences observed in the development of hepatic steatosis could be explained by estrogens, which influence lipid homeostasis, the regulation of adiposity, and liver steatosis. In a mouse model of estrogen insufficiency—the aromatase knockout mouse—hepatic steatosis and age-progressive obesity have been demonstrated. 34, 35 Interestingly, in our study, 25 of the 28 women with moderate/severe steatosis were postmenopausal and over the age of 55, none of them were on hormone replacement therapy. Among women, increasing age and high GGT levels were the only risk factors significantly associated with the degree of steatosis. Among men, IR, advanced fibrosis, and high GGT levels were independent predictors of moderate/severe steatosis. The different pattern of risk factors for steatosis observed between men and women may explain the sex-associated differences in the development of steatosis observed in our cohort.

According to recent reports, central adiposity is an independent predictor of liver steatosis. 36 Stranges et al.36 have demonstrated that abdominal height, a simple anthropometric index of abdominal visceral adipose tissue accumulation, was consistently more correlated to GGT levels than BMI in both sexes. Anthropometrical measurements were not performed in the RCT from which we drew our database. Our data indicate that patients with elevated pretreatment GGT levels have the greatest likelihood of having moderate/severe steatosis. The association between GGT levels and steatosis is likely to result from the association between regional body fat distribution and fatty liver, irrespective of total body fat quantity, which is consistent with the assumption that GGT is a surrogate marker of central fat accumulation. In a cross-sectional study involving 69 randomly selected healthy men aged 38 years, GGT and waist-to-hip ratio were significantly correlated independently from BMI.37 This finding suggests an association between GGT and male pattern (central) obesity, which has emerged as a strong and independent predictor of type 2 diabetes in a number of studies.38, 39 Therefore, GGT level may be a simple and reliable marker of visceral and hepatic fat, and by inference, of hepatic IR. The administration of a questionnaire regarding alcohol intake and drug use or abuse make us confident in excluding any role of these potential confounders on GGT levels. Lack of data on smoking, however, could affect the accuracy of the results.36

In patients with chronic hepatitis C, BMI is closely correlated with the degree of hepatic steatosis. 2, 19 In our study, independent predictor of degree of steatosis included BMI when steatosis was analyzed as a continuous variable. However, BMI was no longer significant when steatosis was represented as a categorical variable. This lack of correlation is probably attributable to the fact that BMI was mostly in the normal range, and to the small number of patients with grades 2 and 3 steatosis in our study population.

Our study provides further evidence that moderate/severe hepatic steatosis is associated with liver fibrosis, indicating that fatty liver, but not IR, is an independent risk factor for advanced fibrosis. Although recent studies have suggested that IR may contribute to the progression of fibrosis in chronic hepatitis C, 15, 32, 40, 41 we could not provide evidence for a direct correlation between IR and severity of fibrosis. Differences in the demographic characteristics of the patients, in the baseline severity of the illness, in the prevalence of genotype 1 infected subjects, and in the number of alcohol abusers among the studies may explain this discrepancy. Considerable variation occurred in the methods for quantifying steatosis and fibrosis on liver biopsy as well as in the cutoffs of steatosis and fibrosis used for statistical analyses among the different studies evaluating the impact of steatosis on fibrosis. This outcome suggests that global standardized criteria for grading and staging histological injury are needed to obtain comparable data.

Our analysis showed that older age, high GGT levels, advanced fibrosis, and moderate/severe steatosis were independent predictors of poor response to antiviral therapy. These findings are in keeping with those of others 21–24 emphasizing the importance of the degree of hepatic fibrosis and steatosis in interferon unresponsiveness. According to recent reports,7, 24 we demonstrate that moderate/severe steatosis is associated with low likelihood of SVR. The obvious clinical implication of this finding is that interventions aimed at preventing or reducing steatosis could attenuate progression of liver fibrosis while increasing the likelihood of responding to antiviral treatment.

Although Romero-Gómez et al. 25 suggested that IR impairs SVR to peginterferon plus ribavirin in 159 patients with chronic hepatitis C (mean age, 41 years; 71% genotype 1), we could not confirm a significant reduction in the response to combination therapy in subjects with IR. Possible explanations are that our cohort included difficult to treat genotype 1 infected patients with a mean age of about 50 years, high baseline viral load, advanced hepatic fibrosis, and mostly with community acquired infection suggesting a long duration of disease.

The main limitation of the current study, as well as of other cross-sectional studies, is that it is unable to distinguish the temporality of the associations between IR, steatosis, and fibrosis. Data by Shintani et al. 16 show that, in a mouse model transgenic for the HCV core gene, signaling abnormalities in the insulin receptor–IRS-1 pathway preceded development of liver steatosis. Lack of data on other potential confounders, such as leptin, adiponectin, or central adiposity could also affect the interpretation of our findings. In addition, we cannot exclude the possibility that denied alcohol abuse may be responsible for the observed prevalence of steatosis. A further methodological issue arises in the potential limitation of the generalizability of results to new populations and settings. Our study included a cohort of nondiabetic European patients who were enrolled in an RCT of peginterferon and ribavirin therapy limiting the broad application of the results.

In conclusion, this study in nondiabetic European patients at low risk for the metabolic syndrome with genotype 1 hepatitis C shows that nearly 60% of subjects had steatosis and that in two thirds of them steatosis was mild. Women, especially when postmenopausal, develop more steatosis than men. Our data support the concept that the causal pathway to steatosis may be different in men and in women, and that IR is a risk factor for moderate/severe steatosis, especially in men. Moderate/severe steatosis has clinical relevance, being associated with severity of fibrosis and hyporesponsiveness to antiviral therapy.

Acknowledgements

The authors thank Giulio Marchesini, M.D., for his help in revising the article, and Warren Blumberg for his patient help in editing the article.

Ancillary