Factors associated with severity of hepatic fibrosis in people with chronic hepatitis C infection

A major feature of hepatitis C virus (HCV) infection is the highly variable course of its natural history. Although most people develop chronic infection, with its consequent risk of cirrhosis, liver failure and hepatocellular carcinoma, only a minority will progress to these advanced liver disease endpoints.1 A key question is whether patients at high risk of disease progression can be distinguished from those with a relatively benign disease course.

Factors previously shown to influence disease progression in chronic hepatitis C have included duration of infection, age at infection, sex, alcohol intake, and co-infection with HIV and hepatitis B virus (HBV).2-6 Evidence about the role of other factors, such as source of HCV infection and ethnicity, is less clear. People with chronic HCV infection and persistently normal serum transaminase levels are less likely to progress to advanced liver disease than those with abnormal levels.7 However, there is conflicting evidence as to the importance of serum transaminase levels in predicting progression among those with abnormal levels.8-10

Consensus guidelines for managing chronic HCV infection generally recommend antiviral therapy for people with moderate to severe hepatic fibrosis, while often recommending observation of people in whom fibrosis is either absent or minimal.11 Thus, correlates of hepatic fibrosis severity could be used to develop assessment and monitoring strategies for chronic HCV infection, including selection of patients for liver biopsy and possible therapeutic intervention.

To assess factors that may influence the development of hepatic fibrosis, we analysed pretreatment demographic and clinical data for patients receiving interferon monotherapy during the mid-1990s in Australia.

Histopathological findings from liver biopsies performed in the six months before baseline assessment were recorded. Portal/periportal and lobular inflammatory activity and staging of fibrosis were generally graded on a five-point (0–4) scoring system according to Scheuer.12 Histopathological classification systems in which the denominator for fibrosis was greater than four were used in some cases, but these patients were excluded from the analyses.

We performed analyses to examine factors which might distinguish patients with moderate to severe hepatic fibrosis (stage 2–3) from those with no or minimal hepatic fibrosis (stage 0–1). The fibrosis scores were grouped in this way, as consensus guidelines now generally recommend antiviral therapy for patients with moderate to severe hepatic fibrosis.11 Patients with stage 4 fibrosis (cirrhosis) were not eligible for S100 interferon therapy, and were excluded.

Seven variables were initially examined among the subgroup of patients with a fibrosis score of 0–3 and a recorded estimated duration of HCV infection. These were age at infection, sex, ethnicity, source of infection, duration of infection, alcohol intake, and mean ALT level. Duration of infection was defined as the time interval between estimated time of exposure and date of liver biopsy. Alcohol intake was estimated from the recorded mean consumption in the previous six months. Mean ALT was calculated from three ALT measurements carried out within six months of the liver biopsy, and expressed as a multiple of ULN for the particular laboratory.

As duration of HCV infection is often difficult to assess, a second analysis was performed to examine factors associated with hepatic fibrosis in the absence of duration of infection. We analysed six variables for all patients with a hepatic fibrosis score of 0–3, whether or not there was an estimated duration of infection. These variables were age at liver biopsy, sex, ethnicity, source of infection, alcohol intake, and mean ALT level.

For both analyses univariate and multivariate models were used to examine associations between selected variables and hepatic fibrosis score. In univariate analyses χ2 tests were used to examine categorical variables. All selected variables were entered into a multivariate logistic regression model to examine independent associations.

Finally, we assessed the relationship between hepatic fibrosis score and both mean ALT level and inflammatory score (portal/periportal, lobular, and combined) using a Wilcoxon-type test for trend.13 Values were considered statistically significant if the P value was < 0.05 or if the 95% confidence intervals did not cross unity.

Current antiviral therapy guidelines strongly recommend therapeutic intervention for patients who have developed moderate to severe hepatic fibrosis,11 as the risk of progression to advanced liver disease is relatively high in this group.14 Our study suggests that, in chronic HCV infection, more severe hepatic fibrosis is associated with older age at infection, longer duration of infection, alcohol intake, and higher ALT levels. In contrast, sex, ethnicity and source of infection did not influence the extent of hepatic fibrosis.

There were several limitations in our methods. First was the use of cross-sectional data to examine a longitudinal process. Cross-sectional analyses rely on estimation of time of HCV exposure, which is often problematic. To help counter this, we analysed separately patients in whom hepatic fibrosis score and time of HCV exposure were recorded, and all patients with hepatic fibrosis score with no regard to time of HCV exposure. Secondly, a large proportion of patients did not have a hepatic fibrosis score recorded, and there were differences in demographic and clinical characteristics between patients with and without a fibrosis score. Thirdly, being based on liver clinic populations, the participants were probably not totally representative of the Australian population with hepatitis C. Finally, different histopathological classification systems for liver biopsy grading and staging were used, but final analyses were limited to patients whose fibrosis was staged under a five-point scoring system.

A link between age at HCV infection and the presence of hepatic fibrosis is supported by the strong and consistent trend of increasing risk with age at infection, and the independence of effect. This relationship has been seen in several previous cross-sectional studies.2,15,16 Some recent longitudinal studies have found very low rates of progression to advanced liver disease among young adults infected through injecting drug use and contaminated anti-D immunoglobulin injections,17-19 and among children infected through blood transfusion.20 This association may also partly explain the relatively more rapid disease progression in adult post-transfusion cohorts,21-24 as the median age at infection in these studies was generally 40–50 years.

We also found that duration of chronic HCV infection was independently associated with presence of hepatic fibrosis. Again, such a finding is consistent with other studies.1,2 The cross-sectional nature of our study precludes examination of the possible non-linearity of disease progression, recently hypothesised by others.25

Our study shows that age alone is an important predictor of hepatic fibrosis score. The risk of moderate to severe hepatic fibrosis increased significantly after 35 years of age in our study population. This finding suggests that older patients with chronic HCV infection and abnormal ALT levels who are unable to identify their time of HCV exposure (and therefore their duration of infection) should strongly consider liver-biopsy staging of disease.

Alcohol intake is often difficult to estimate accurately, and consumption may be modified after HCV infection is diagnosed. Current alcohol intake of less than 70 g per week was also a requirement for access to interferon therapy. In our study, alcohol intake in the previous six months was not reported for many patients, and only a small percentage (3%) gave a history of heavy intake (> 40 g/day). Despite this, alcohol intake was associated with the presence of hepatic fibrosis, as several other studies have noted,2,26,27 although the increase in the odds ratio was relatively small.

The relationship between serum transaminase levels and risk of liver disease progression remains controversial. Although risk of disease progression is considerably higher among people with abnormal serum transaminase levels compared with those with consistently normal levels,7 there is conflicting evidence on the association between degree of ALT abnormality and extent of histological fibrosis. Several studies have shown either no correlation8,28,29 or a very weak correlation30,31 between ALT level and hepatic fibrosis. However, these studies have had relatively small study populations, with fewer than 100 patients undergoing liver biopsy. Our findings suggest a "dose–response" pattern for this relationship.

The pathogenesis of hepatic fibrosis in chronic HCV infection is not well understood; however, an association between inflammatory activity and fibrogenesis is likely. Progression to cirrhosis appears to occur more commonly in patients with chronic HCV infection and high-grade necroinflammatory activity on liver biopsy.32 Our findings suggest that portal/periportal inflammatory activity may play a more important role in fibrogenesis than lobular activity. This finding is consistent with the fact that the initial focus of hepatic fibrosis is in the portal region.

This study adds to previous Australian reports14,33,34 and contributes to a clearer picture of the natural history of chronic hepatitis C, particularly in the Australian context. As patients with moderate to severe hepatic fibrosis have a greatly increased risk of progression to advanced liver disease compared with those with earlier stages of disease,14 the factors that we have identified should improve our ability to predict patients with progressive disease. Patients with higher ALT levels, longer duration of infection, older age, and those who drink alcohol warrant special attention, with more intensive follow-up. These factors may also assist in targeting antiviral therapeutic intervention.