Patients and methods

Patient population

This prospective study included 226 consecutive patients with HCV-4 who underwent liver biopsy in our department between January 2004 and December 2008. The following conditions were excluded: positive hepatitis B surface antigen, human immunodeficiency virus infection, autoimmune hepatitis, haemochromatosis, α1-antitrypsin deficiency, Wilson’s disease, and hepatic decompensation. All patients gave their informed consent prior to liver biopsy.

Clinical and laboratory assessment

The following data were collected at the time of liver biopsy: gender, age, geographical origin, daily alcohol intake in the past 6 months (g/day), route of HCV transmission, body mass index (BMI), and routine laboratory tests (complete blood count, biochemical tests). Serum insulin and C-peptide were determined by electrochemiluminescence immunoassay (Elecsys 2010; Roche Diagnostics, Indianapolis, Indiana, USA) after an overnight fast of 12 h. IR was assessed using the homeostasis model (HOMA-IR), and arbitrarily defined as HOMA-IR >3.

Virological assessment

HCV subtypes were determined using the Versant HCV genotype 2.0 (LiPA) assay (Siemens Medical Solutions, Puteaux, France). Serum HCV-RNA was quantified using the Versant HCV-RNA 3.0 (bDNA) Assay (Siemens Medical Solutions). High viral load was defined as serum HCV-RNA >800 000 IU/ml. During the treatment course, serum HCV-RNA level was measured at treatment weeks 12, 24 and 48, and 24 weeks after the end of treatment. Serum HCV-RNA was assessed using the Versant HCV-RNA Qualitative (TMA) Assay (Siemens Medical Solutions) for samples undetectable by bDNA.22

Treatment

The benefit/risks of peginterferon plus ribavirin therapy were explained to the whole study population, but patients were free to undergo treatment since this was not a clinical trial. Thus, 108 patients received a 48 week course of peginterferon alfa-2a (180 μg/week) or peginterferon alfa-2b (1.5 μg/kg/week) plus ribavirin (1000 mg or 1200 mg/day for body weight ⩽ or >75 kg respectively) during the study period. Early virological response (EVR) was defined as ⩾2 log decline in serum HCV-RNA level from baseline or undetectable HCV-RNA in serum at treatment week 12. Treatment was stopped at week 24 if serum HCV-RNA was still detectable (TMA). End of treatment (EOT) response was defined as undetectable serum HCV-RNA at the end of treatment. SVR was defined as undetectable serum HCV-RNA 24 weeks after the end of treatment.

Liver histology

Liver biopsy specimens were analysed by a single experienced pathologist who was blinded to clinical and biological data. Necroinflammation and fibrosis were assessed using the METAVIR score.23 Necroinflammation activity (A) was graded as A0 (absent), A1 (mild), A2 (moderate), or A3 (severe). Fibrosis stage (F) was scored as F0 (absent), F1 (portal fibrosis), F2 (portal fibrosis with few septa), F3 (septal fibrosis), and F4 (cirrhosis). Severe fibrosis was defined as METAVIR score F3–F4. Steatosis was assessed as the percentage of hepatocytes containing macrovesicular fat droplets. It was graded as: 0 (absent (<5%)), 1 (moderate (5–30%)), and 2 (severe (>30%)).

Statistical analyses

Continuous variables were summarised as mean (with the SD) and categorical variables as frequency and percentage. Comparisons between groups were performed using the Student t test or the Mann–Whitney U test for continuous variables, and the χ² test or Fisher’s exact test for categorical data. Comparisons between different groups were performed using analysis of variance (standard or non-parametric as appropriate). Correlations between pairs of numerical variables were performed using Spearman’s rank correlation method. All p values were based on a two-sided test of statistical significance. Significance was accepted at p<0.05. To eliminate a potential bias related to the use of an arbitrary cut-off of HOMA-IR to define insulin resistance, the analyses were repeated using HOMA-IR as a continuous variable.

Multiple logistic regression including epidemiological, metabolic, virological and histological features, was used to identify independent factors associated with insulin resistance, severe fibrosis and SVR in the whole study population, as well as in subgroup analyses according to the geographical origin of patients. Variables were described as odds ratio (OR) with 95% confidence interval (CI). All analyses were performed with SPSS software for Windows, version 13.0.

Results

Patient population

The characteristics of the 226 patients with HCV-4 at the time of liver biopsy are shown in table 1. Male gender was predominant (74%). Mean age was 45 (SD 8) years, >45 years in 48% of patients. Ninety-two patients were Egyptian (40%), 80 were European (35%) and 54 were African (24%). Fourteen patients (6%) had excessive alcohol intake (>30 g/day). Intravenous drug use was the main route of HCV transmission in Europeans (41%), while the majority of Egyptians reported a history of parenteral treatment for Schistosoma mansoni infection during childhood or adolescence. Mean BMI was 26.0 (SD 3.8) kg/m², and 52% of patients were overweight (BMI >25 kg/m²). Mean HOMA-IR was 3.7 (SD 4.0), >3 in 46% of patients. Mean serum alanine aminotransferase (ALT) level was 100 (SD 92) IU/l (normal value <40 IU/l). The distribution of HCV-4 subtypes was: a (54%), d (18%) and e/f/g/h/i (27%). Mean serum HCV-RNA level was 5.3 (SD 0.8) log₁₀ IU/ml, >800 000 IU/ml in 30% of patients. Liver histology showed moderate–severe necroinflammation (METAVIR grade A2–A3) in 66 patients (29%), severe fibrosis (METAVIR score F3–F4) in 67 patients (29%), and severe steatosis (>30%) in 58 patients (26%).

Geographical origin

The characteristics of patients differed significantly according to their geographical origin (table 1). Egyptians were significantly younger than Europeans and Africans (43 (SD 7), 45 (SD 7), and 48 (SD 9) years respectively, p<0.001). Intravenous drug use was the main route of HCV transmission in Europeans (41%) compared to Egyptians (2%) and Africans (2%), p = 0.008. Europeans had significantly lower BMI than Egyptians and Africans (24.7 (SD 3.3), 26.7 (SD 3.9), and 26.5 (SD 3.7) kg/m² respectively, p = 0.001). Severe fibrosis (METAVIR score F3–F4) and severe steatosis (>30%) were significantly less frequent in Africans (11% and 5%, respectively) compared to Egyptians (43% and 30%, respectively) and Europeans (26% and 34%, respectively), p<0.001. By contrast, serum ALT level, HOMA-IR, serum HCV-RNA level and liver necroinflammation activity were not significantly different between Egyptians, Europeans and Africans (table 1).

Insulin resistance

Insulin resistance (HOMA-IR >3) was present in 105 patients among the 226 patients with HCV-4 (46%). By univariate analyses (table 2), insulin resistance was associated with: age >45 years (61/105, 58% vs 47/121, 39%, respectively, p = 0.003), BMI >25 kg/m² (68/105, 64% vs 51/121, 42%, respectively, p<0.001), high serum ALT levels (120 (SD 75) vs 88 (SD 108) IU/l, p = 0.021), serum HCV-RNA >800 000 IU/ml (43/105, 41% vs 24/121, 20%, respectively, p<0.001), severe fibrosis (METAVIR score F3–F4) (44/105, 42% vs 23/121, 19%, respectively, p<0.001), and severe steatosis (>30%) (39/105, 37% vs 19/121, 15%, respectively, p<0.001). By contrast, insulin resistance was not associated with geographical origin, gender, alcohol intake, route of HCV transmission, subtypes of HCV-4, and liver necroinflammation activity (table 2). By logistic regression (table 3), insulin resistance was independently associated with: age >45 years (p = 0.006; OR, 2.614; CI, 1.316 to 5.194), BMI >25 kg/m² (p = 0.036; OR, 2.105; CI, 1.048 to 4.229), serum HCV-RNA >800 000 IU/ml (p = 0.002; OR, 3.143; 95% CI, 1.503 to 6.574), severe fibrosis (p = 0.015; OR, 2.657; 95% CI, 1.214 to 5.818), and severe steatosis (p = 0.028; OR, 2.488; 95% CI, 1.105 to 5.602).

Insulin resistance was present in 28 patients among the 97 patients (29%) who were not overweight (BMI <25 kg/m²) and had no cirrhosis (METAVIR score <F4), and was associated with serum HCV-RNA >800 000 IU/ml uniquely (15/33, 45% vs 13/64, 20%, respectively, p = 0.009). In addition, HOMA-IR correlated significantly with serum HCV-RNA level in these patients (Spearman rank correlation 0.323, p = 0.002) (fig 1). It is noteworthy that insulin resistance was not associated with geographical origin of these patients, a result to be considered cautiously given the low number of patients in each group.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Serum HCV-RNA level (mean with the SD) in patients with insulin resistance (HOMA-IR>3) compared to those without insulin resistance (HOMA-IR <3). HCV, hepatitis C virus; HOMA-IR, homeostasis model assessment of insulin resistance.

Liver fibrosis

Liver biopsy was performed for all the 226 patients with HCV-4. The mean length of liver biopsies was 16 (SD 6) mm (range, 10–35 mm). Moderate–severe necroinflammation (METAVIR grade A2–A3) was present in 66 patients (29%). Severe fibrosis (METAVIR score F3–F4) was present in 67 patients (29%). Severe steatosis (>30%) was present in 58 patients (26%). By univariate analyses (table 2), severe fibrosis was associated with Egyptian origin (40/67, 59% vs 52/159, 32%, respectively, p<0.001), male gender (56/67, 83% vs 112/159, 70%, respectively, p = 0.039), excessive alcohol intake (8/67, 12% vs 6/159, 3%, respectively, p = 0.030), HCV-4 subtype a (45/67, 67% vs 78/159, 49%, respectively, p = 0.013), HOMA-IR >3 (42/67, 62% vs 63/159, 39%, respectively, p = 0.001), and severe steatosis (28/67, 41% vs 30/159, 19%, respectively, p<0.001). By contrast, severe fibrosis was not associated with age, BMI, route of HCV transmission, serum ALT level, serum HCV-RNA level, and liver necroinflammation activity (table 2). By logistic regression, severe fibrosis was independently associated with Egyptian origin (p<0.001; OR, 5.872; 95% CI, 2.747 to 12.553), excessive alcohol intake (p = 0.021; OR, 5.311; 95% CI, 1.287 to 21.924), and HOMA-IR >3 (p<0.001; OR, 3.864; 95% CI, 1.859 to 8.034) (table4). It is noteworthy that HOMA-IR remained an independent predictor of severe fibrosis when used as a continuous variable to define insulin resistance (p = 0.003; OR, 1.144; CI, 1.047 to 1.250). Interestingly, the analyses were repeated after the exclusion of the 21 patients with cirrhosis (METAVIR score F4) and confirmed that Egyptian origin, excessive alcohol intake, and insulin resistance were the independent predictors of severe fibrosis (METAVIR score F3) in patients with HCV-4 (results not shown).

Severe fibrosis was present in 40 patients among the 92 Egyptians (43%) (fig 2) and was associated with older age (45 (SD 6) vs 41 (SD 7) respectively, p = 0.032), high serum ALT levels (112 (SD 72) vs 80 (SD 56) respectively, p = 0.029), and HOMA-IR >3 (23/40, 57% vs 16/52, 30% respectively, p = 0.010). By logistic regression, severe fibrosis was independently associated with HOMA-IR >3 uniquely (p = 0.009; OR, 3.516; 95% CI, 1.375 to 8.994). By contrast, severe fibrosis was present in only 27 patients among the 134 non-Egyptians (20%) (80 Europeans and 54 Africans; fig 2) and was associated with excessive alcohol intake (7/27, 26% vs 6/107, 5% respectively, p = 0.004), HOMA-IR >3 (20/27, 74% vs 46/107, 43%, respectively, p = 0.003) and severe steatosis (15/27, 55% vs 15/107, 14%, respectively, p<0.001). By logistic regression, severe fibrosis was independently associated with excessive alcohol intake (p = 0.037; OR, 4.808; 95% CI, 1.095 to 21.103), HOMA-IR >3 (p = 0.045; OR, 3.132; 95% CI, 1.023 to 9.585), and severe steatosis (p = 0.004; OR, 4.940; 95% CI, 1.679 to 14.528).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Percentage of patients with insulin resistance (HOMA-IR >3), severe fibrosis (METAVIR score F3–F4), and sustained virological response (SVR) in patients with hepatitis C virus genotype 4 (HCV-4), according to geographical origin.

Sustained virological response

One hundred and eight patients out of the 226 patients with HCV-4 received a 48 week course of peginterferon and ribavirin during the study period. Fifty-two patients (48%) were Egyptian, 33 (30%) were European, and 23 (21%) were African. Sixty-seven patients (62%) received peginterferon alfa-2a plus ribavirin, and 41 patients (38%) received peginterferon alfa-2b plus ribavirin. Eighty patients (74%) developed EVR, among whom 44 had undetectable serum HCV-RNA (40%), while the remaining 36 (33%) showed a decline >2 log₁₀ IU/ml of serum HCV-RNA from baseline levels. Sixty-nine patients (64%) developed EOT response but only 59 (54%) achieved SVR, while 10 (9%) relapsed during follow-up. Among the 80 patients with EVR, 58 achieved SVR (positive predictive value 72%). By contrast, 27 patients out of the 28 with no EVR failed treatment (negative predictive value 96%). Stepwise reductions in the doses of peginterferon and/or ribavirin were necessary in 19 patients (17%) for haematological side effects, but without significant difference between patients who achieved SVR (8/59, 13%) and those who did not (11/49, 22%), p = 0.227. Moreover, there was no treatment discontinuation for adverse events.

By univariate analyses (table 5), SVR was associated with: Egyptian origin (33/59, 56% vs 19/49, 38%, respectively, p = 0.047), younger age (43 (SD 7) vs 46 (SD 7) respectively, p = 0.026), HOMA-IR <2 (39/59, 66% vs 14/49, 28% respectively, p<0.001), serum HCV-RNA <800 000 IU/ml (45/59, 76% vs 27/49, 55%, respectively, p = 0.020), and non-severe fibrosis (METAVIR score F0–F2) (47/59, 79% vs 24/49, 49%, respectively, p<0.001). Interestingly, EVR was also associated with HOMA-IR <2 (47/80, 58% vs 6/28, 21%, respectively, p<0.001).

By contrast, SVR was not associated with gender, alcohol intake, route of HCV transmission, BMI, subtypes of HCV-4, treatment formulation, liver necroinflammation activity, and liver steatosis (table 5).

By logistic regression (table 6), SVR was independently associated with Egyptian origin (p<0.001; OR, 13.119; 95% CI, 3.089 to 55.706), HOMA-IR <2 (p = 0.001; OR, 5.314; 95% CI, 1.953 to 14.459), and non-severe fibrosis (p<0.001; OR, 8.059; 95% CI, 2.512 to 25.855). It is noteworthy that HOMA-IR remained an independent predictor of SVR when used as a continuous variable to define insulin resistance (p = 0.030; OR, 0.848; CI, 0.731 to 0.984).

SVR was achieved in 33 patients among the 52 Egyptians (63%) who received peginterferon and ribavirin therapy (fig 2). By univariate analyses SVR was associated with HOMA-IR <2 (18/33, 54% vs 5/19, 26%, respectively, p = 0.048) and non-severe fibrosis (23/33, 69% vs 4/19, 21% respectively, p = 0.001). By logistic regression, SVR was independently associated with both HOMA-IR <2 (p = 0.049; OR, 5.850; 95% CI, 1.010 to 33.872) and non-severe fibrosis (p = 0.019; OR, 5.998; 95% CI, 1.345 to 26.742). By contrast, SVR was achieved only in 26 patients among the 56 non-Egyptians (46%) who received peginterferon and ribavirin therapy (33 Europeans and 23 Africans; fig 2) and was associated with younger age (43 (SD 8) vs 48 (SD 7) years respectively, p = 0.029), serum HCV-RNA <800 000 IU/ml (22/26, 84% vs 14/30, 46%, respectively, p = 0.003), HOMA-IR <2 (21/26, 80% vs 9/30, 30%, respectively, p<0.001) and non-severe fibrosis (24/26, 92% vs 20/30, 66%, respectively, p = 0.020). By logistic regression, SVR was independently associated with HOMA-IR <2 uniquely (p = 0.001; OR, 10.000; 95% CI, 2.621 to 38.149).

Discussion

This study provides complete information about the epidemiological, metabolic, virological, histological and therapeutic features of HCV-4 in a large prospective cohort of patients followed in France. Importantly, the three major geographical origins of HCV-4 were represented in this cohort, including patients from Egypt, Europe and Africa. The first finding of this study was the heterogeneity of the HCV-4 disease according to the geographical origin of patients (table 1). In this respect, Egyptians were significantly younger than Europeans and Africans, contracted their HCV mainly through parenteral treatment for schistosomiasis, were infected in the vast majority of cases with the subtype 4a, and presented frequently with severe fibrosis. In contrast, Africans were significantly older than Egyptians and Caucasians, were infected with various subtypes of the virus, and presented frequently with non-severe fibrosis. As for the Europeans, they were thinner than Egyptians and Africans, contracted their HCV mainly through intravenous drug use, were infected mainly by the subtypes 4a and 4d of the virus, and presented with a severe fibrosis in a significant proportion of cases. Importantly, insulin resistance, assessed for the first time in a large cohort of HCV-4 patients, was found to be highly prevalent with 46% of patients showing HOMA-IR >3. It is noteworthy that there was no significant difference in the rate of insulin resistance according to the geographical origin of patients (fig 2). This finding may be related to the heterogeneous features of the different groups especially age, BMI, and liver fibrosis, the three major determinants of insulin resistance in HCV patients (table 2).24 25 Interestingly, insulin resistance was associated in multivariate analysis with these classical variables, but the strongest independent predictor was the high viral load (serum HCV-RNA >800 000 IU/ml) suggesting a direct role of the virus in the development of insulin resistance (table 3).26 27 28 29 30 This hypothesis was reinforced by finding a considerable proportion of patients who showed insulin resistance despite the absence of overweight and cirrhosis (nearly 30%), and the significant correlation of HOMA-IR with serum HCV-RNA levels in these patients (fig 1).

The second finding of this study was the predictors of severe fibrosis in HCV-4 patients (table 4). In this respect, Egyptian origin was independently associated with severe fibrosis. This association should be considered with care. In fact, severe fibrosis may be related to a longer duration of the disease since most patients contracted HCV during childhood or adolescence by parenteral treatment of schistosomiasis. Unfortunately, the duration of the disease was not entered as a variable in the statistical analyses since the route (and subsequently the date) of transmission were not determined for the majority of Europeans and Africans. Insulin resistance was also independently associated with severe fibrosis.31 This is consistent with previous studies in other genotypes,6 7 8 9 10 11 12 13 and was demonstrated for the first time for HCV-4 patients in this study. Interestingly, insulin resistance remained a major predictor of severe fibrosis even when patients with cirrhosis were excluded from the analyses, cirrhosis being a potential confounding factor influencing the insulin resistance status. Given the fact that severe fibrosis was most frequent in Egyptians (fig 2), we also performed a subgroup analysis based on the geographical origin of patients, and found that insulin resistance remained a major predictor of severe fibrosis in both groups (Egyptians and non-Egyptians). It is of note that severe fibrosis was associated with steatosis in univariate analyses but not in logistic regression. Given the cross-sectional type of our study, results should be interpreted as statistical associations rather than pathophysiological findings. In fact the relationship between steatosis, insulin resistance and liver fibrosis in the context of chronic hepatitis C is very complex. Longitudinal studies with serial liver biopsies are definitively needed to further characterise and understand these mechanisms.

The third finding of this study was the predictors of SVR in patients with HCV-4 treated with peginterferon plus ribavirin. In this respect, Egyptian origin was independently associated with SVR despite the presence of severe fibrosis in a large proportion of patients, a strong factor with negative influence on SVR rates. The reason of this finding is not clear. Two hypotheses may be generated. The first is based on a better susceptibility of subtype 4a (the major subtype in Egypt) to antiviral treatment; in this respect, despite the absence of significant association between SVR and HCV-4 subtypes (table 5), SVR rates were significantly higher in Egyptian versus European and African subtype 4a (61% vs 45% vs 16%, respectively). The second hypothesis is based on a different genetic background and subsequent different immune response, as previously implicated to explain the disparity in response rates between Caucasians and Africans.32 33 In fact, the difference in SVR rates between Egyptians and non-Egyptians could not be explained by factors known to be associated with poor response such as baseline HCV-RNA levels, sex, age, weight, degree of fibrosis, or amount of drug taken. These results should be considered cautiously given the relatively low number of treated patients in each group. Randomised clinical trials including a large number of patients in each group are definitively needed in the future to better elucidate the difference observed in SVR rates in HCV-4 patients.

Interestingly, HOMA-IR <2, the unique predictor of SVR found in Egyptians, was also independently associated with SVR in non-Egyptians as well as in all 108 treated patients irrespectively of their geographical origin. This effect appears to be independent of fibrosis stage, is consistent with previous studies in other genotypes,18 19 20 21 and, in this study, was demonstrated for the first time for patients with HCV-4. Therapeutic intervention aimed at decreasing insulin resistance may therefore be warranted in these patients before or during peginterferon and ribavirin treatment. In this respect, a recent study in patients with the HCV genotype 1 and with HOMA-IR >2 showed a better SVR rate in patients treated peginterferon and ribavirin in association with metformin in comparison to those who did not receive metformin (52% vs 42%).34

In conclusion, insulin resistance and geographical origin are major predictors of liver fibrosis and response to peginterferon and ribavirin therapy in patients with HCV-4. Insulin resistance is frequently encountered in such patients and is independently correlated with viral load.