Hepatitis C Therapy in Elderly Patients
Hepatitis C Therapy in Elderly Patients
The Swiss Hepatitis C Cohort Study (SCCS) enrols HCV patients >18 years at 8 secondary and tertiary healthcare centres since the year 2000. For this study, data from all SCCS patients who had received antiviral therapy with PEG-IFN-α and ribavirin were analysed. Patients <60 years at the time of therapy were defined as 'young patients' and patients ≥60 years were defined as 'elderly patients', as done in other publications on this topic. As other studies defined elderly as ≥65 years, a subgroup analysis was performed for patients ≥65 years. Patients with hepatitis B or human immunodeficiency virus coinfection and current alcohol consumption >40 g/day were excluded.
Based on these criteria, 516 patients could be analysed, including 66 elderly patients. Data about demographic factors (age, gender, race), baseline virological factors (genotype, viral load), host factors and comorbidity (body mass index, fibrosis stage and presence of cirrhosis, haemoglobin level, platelet count, previous HCV therapy), treatment factors (doses of PEG-IFN-α and ribavirin, duration of treatment), SVR and clinical outcome (death) were retrieved from the SCCS database. Liver fibrosis stage was classified according to the Metavir score by expert pathologists at each study centre, with biopsy results available in 76.7% of patients. In patients without liver biopsy (23.3% of patients), the presence or absence of cirrhosis was assumed based on predefined criteria including signs of decompensated liver disease such as gastrointestinal bleeding, jaundice, hepatic encephalopathy and ascites.
PEG-IFN-α and ribavirin treatment duration was 48 weeks for genotypes 1/4/6 and 24 weeks for genotypes 2/3. Discontinuation of treatment was defined as any treatment duration of <44 weeks in genotypes 1/4/6 and <20 weeks in genotypes 2/3. Furthermore, we looked for dose modifications of ribavirin and PEG-IFN-α. Elderly patients had been more frequently pretreated than younger patients (24% vs. 12.4%). However, previous treatment was performed with PEG-IFN-α and ribavirin only in six cases (7.6%).
Descriptive statistics of baseline characteristics and treatment outcome were performed with mean ± standard deviation for the continuous variables, and with percentages for the categorical variables. We applied t-tests and chi-squared tests for group comparisons. A logistic regression model was used to assess whether age has an independent influence on the primary endpoint, SVR per therapy course, considering gender, cirrhosis, HCV genotype, previous HCV therapy and viral load (defined as >600 000 IU/ml) as confounders. Generalized estimating equations (GEE) were applied in this context to account for the fact that not all observations on treatment success are independent in the sense that we had some patients observed 2–3 times in the database. Secondly, a matched pair analysis was performed. For each elderly patient, we identified a younger patient with identical set of confounders (gender, cirrhosis, HCV genotype, previous HCV therapy and viral load). In the case when more than one control patient could be matched with an elderly patient, a control patient was randomly drawn. We then estimated the proportion of SVR in both groups and compared them with a binomial test. In a third analysis, classification trees were used to predict SVR, with age and the five confounders described above as explanatory variables. The significance level was set to 5%. All analyses were performed with R 2.13.
Materials and Methods
Patients
The Swiss Hepatitis C Cohort Study (SCCS) enrols HCV patients >18 years at 8 secondary and tertiary healthcare centres since the year 2000. For this study, data from all SCCS patients who had received antiviral therapy with PEG-IFN-α and ribavirin were analysed. Patients <60 years at the time of therapy were defined as 'young patients' and patients ≥60 years were defined as 'elderly patients', as done in other publications on this topic. As other studies defined elderly as ≥65 years, a subgroup analysis was performed for patients ≥65 years. Patients with hepatitis B or human immunodeficiency virus coinfection and current alcohol consumption >40 g/day were excluded.
Based on these criteria, 516 patients could be analysed, including 66 elderly patients. Data about demographic factors (age, gender, race), baseline virological factors (genotype, viral load), host factors and comorbidity (body mass index, fibrosis stage and presence of cirrhosis, haemoglobin level, platelet count, previous HCV therapy), treatment factors (doses of PEG-IFN-α and ribavirin, duration of treatment), SVR and clinical outcome (death) were retrieved from the SCCS database. Liver fibrosis stage was classified according to the Metavir score by expert pathologists at each study centre, with biopsy results available in 76.7% of patients. In patients without liver biopsy (23.3% of patients), the presence or absence of cirrhosis was assumed based on predefined criteria including signs of decompensated liver disease such as gastrointestinal bleeding, jaundice, hepatic encephalopathy and ascites.
Treatment Including Dose Reduction and Discontinuation
PEG-IFN-α and ribavirin treatment duration was 48 weeks for genotypes 1/4/6 and 24 weeks for genotypes 2/3. Discontinuation of treatment was defined as any treatment duration of <44 weeks in genotypes 1/4/6 and <20 weeks in genotypes 2/3. Furthermore, we looked for dose modifications of ribavirin and PEG-IFN-α. Elderly patients had been more frequently pretreated than younger patients (24% vs. 12.4%). However, previous treatment was performed with PEG-IFN-α and ribavirin only in six cases (7.6%).
Statistical Analyses
Descriptive statistics of baseline characteristics and treatment outcome were performed with mean ± standard deviation for the continuous variables, and with percentages for the categorical variables. We applied t-tests and chi-squared tests for group comparisons. A logistic regression model was used to assess whether age has an independent influence on the primary endpoint, SVR per therapy course, considering gender, cirrhosis, HCV genotype, previous HCV therapy and viral load (defined as >600 000 IU/ml) as confounders. Generalized estimating equations (GEE) were applied in this context to account for the fact that not all observations on treatment success are independent in the sense that we had some patients observed 2–3 times in the database. Secondly, a matched pair analysis was performed. For each elderly patient, we identified a younger patient with identical set of confounders (gender, cirrhosis, HCV genotype, previous HCV therapy and viral load). In the case when more than one control patient could be matched with an elderly patient, a control patient was randomly drawn. We then estimated the proportion of SVR in both groups and compared them with a binomial test. In a third analysis, classification trees were used to predict SVR, with age and the five confounders described above as explanatory variables. The significance level was set to 5%. All analyses were performed with R 2.13.
Source...