Peginterferon alpha-2b is safe and effective in HBeAg-positive chronic hepatitis B patients with advanced fibrosis
Hepatology July 2007
Erik H. C. J. Buster 1, Bettina E. Hansen 1 2, Maria Buti 3, Jean Delwaide 4, Claus Niederau 5, Peter P. Michielsen 6, Robert Flisiak 7, Pieter E. Zondervan 8, Solko W. Schalm 1, Harry L. A. Janssen 1 *, HBV 99-01 study group
1Department of Gastroenterology and Hepatology, Erasmus MC University Medical Center Rotterdam, Rotterdam, the Netherlands
2Department of Epidemiology and Biostatistics, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands
3Liver Unit, Hospital General Universitari Pall d'Hebron, Barcelona, Spain
4Department of Hepato-Gastroenterology, CHU Sart Tilman, Liege, Belgium
5Department of Hepatology, Heinrich-Heine-Universitat, Dusseldorf, Germany
6Department of Gastroenterology and Hepatology, University Hospital, Antwerp, Belgium
7Department of Infectious Diseases, Medical University of Biaystok, Poland
8Department of Pathology, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands
Schering-Plough International, Kenilworth, NJ
GlaxoSmithKline, Research and Development, Greenford, UK
Foundation for Liver Research (SLO), Rotterdam, The Netherlands
Chronic hepatitis B (CHB) patients with advanced fibrosis are often not considered for treatment with peginterferon (PEG-IFN) because IFN therapy may precipitate immunological flares, potentially inducing hepatic decompensation.
We investigated the efficacy and safety of treating hepatitis B e antigen (HBeAg)-positive CHB patients with 52 weeks of PEG-IFN-a-2b (100ug weekly) alone or in combination with lamivudine (100 mg daily). Seventy patients with advanced fibrosis (Ishak fibrosis score 4-6) and 169 patients without advanced fibrosis, all with compensated liver disease, participated in the study.
Virologic response, defined as HBeAg seroconversion and hepatitis B virus (HBV) DNA < 10,000 copies/ml at week 78, occurred significantly more often in patients with advanced fibrosis than in those without (25% versus 12%, respectively; P = 0.02). Also patients with cirrhosis (n = 24) exhibited a virologic response more frequently than did patients without cirrhosis (30% versus 14%, respectively; P = 0.02).
Improvement in liver fibrosis occurred more frequently in patients with advanced fibrosis (66% versus 26%, P < 0.001). HBV genotype A was more prevalent among patients with advanced fibrosis than among those without (57% versus 24%, P < 0.001).
Most adverse events, including serious adverse events, were observed equally as frequently in patients with advanced fibrosis and those without. Fatigue, anorexia, and thrombocytopenia occurred more often in patients with advanced fibrosis than in those without (P < 0.01). Necessary dose reduction or discontinuation of therapy was comparable for both patient groups (P = 0.92 and P = 0.47, respectively).
Conclusion: PEG-IFN is effective and safe for HBeAg-positive patients with advanced fibrosis. Because PEG-IFN therapy results in a high rate of sustained off-therapy response, patients with advanced fibrosis or cirrhosis but compensated liver disease should not be excluded from PEG-IFN treatment.
Hepatitis B virus (HBV) is one of the most prevalent viral pathogens in humans, with almost a third of the world's population having evidence of infection with HBV and about 400 million people chronically infected. Chronic HBV is a leading cause of cirrhosis and hepatocellular carcinoma (HCC). The most important objective in treatment of chronic HBV is to halt progression of liver injury by persistent suppression of the virus. Complete response, with loss of hepatitis B surface antigen (HBsAg) and appearance of anti-HBs, is difficult to achieve with currently available antiviral drugs and most likely does not mean complete eradication but rather control of residual virus by a persistent host immune response. Therefore, surrogate markers of short-term treatment response, such as loss of hepatitis B e antigen (HBeAg) and decrease in HBV DNA, are most often used to evaluate the effectiveness of antiviral therapy.
Treatment with interferon (IFN) results in loss of HBeAg in about one-third of patients. Somewhat higher response rates have been reported for peginterferon (PEG-IFN) therapy in hepatitis B e antigen (HBeAg)-positive chronic HBV patients.[4-6] HBeAg loss after IFN-based therapy is durable for most patients and has been shown to result in decreased progression of fibrosis, reduced risk of developing hepatocellular carcinoma, and improved survival. Persistence of HBeAg, on the other hand, results in an unfavorable long-term outcome. In 20%-40% of patients, IFN induces immune-mediated hepatitis flares that may precipitate hepatic decompensation in those with advanced liver disease.
Studies of treating chronic HBV patients with conventional IFN or PEG-IFN have identified similar predictors of response. These factors included high pretreatment alanine aminotransferase (ALT) level, low pretreatment HBV DNA, high degree of necroinflammatory activity, infection as an adult, and HBV genotype A or B.[13-16] PEG-IFN often is not given to HBV-infected patients with advanced fibrosis or cirrhosis because of its presumed lack of efficacy and the fear that hepatic flares and other toxicity-related events will occur.
In this study we investigated the response to and safety of PEG-IFN alpha-2b (PEG- IFN-a-2b) alone or in combination with lamivudine in patients with HBeAg-positive chronic hepatitis B and advanced fibrosis.
Patients and Methods
We investigated the efficacy and safety of 52 weeks of PEG-IFN-a-2b alone or in combination with lamivudine in patients with HBeAg-positive chronic HBV and advanced fibrosis. The patients participating and the results of the multicenter, randomized, controlled study investigating PEG-IFN-a-2b monotherapy and PEG-IFN--2b in combination with lamivudine in those with HBeAg-positive chronic HBV were described previously. All patients had been positive for hepatitis B s antigen (HBsAg) for at least 6 months prior to randomization. Patients were eligible if they tested HBeAg positive on 2 occasions less than 8 weeks prior to randomization, had a serum ALT level that was more than 2 up to 10 times the upper limit of normal (ULN), and had serum HBV DNA greater than 1.0 X 105 copies/ml. Major exclusion criteria were: antiviral therapy less than 6 months prior to randomization; serum antibodies to hepatitis C virus, hepatitis D virus, or human immunodeficiency virus; preexisting leukopenia (white blood cell count <3,000/mm3, neutrophil count <1,800/mm3) or thrombocytopenia (platelets 100,000/mm3); or decompensated liver disease. Patients were randomized at a 1:1 ratio to receive PEG-IFN-a-2b (100 ug weekly) with placebo or with lamivudine (100 mg daily) for 52 weeks. After 32 weeks, PEG-IFN-a-2b dosage was lowered to 50 ug per week to prevent side effects and early treatment discontinuation. Follow-up after discontinuation of therapy lasted 26 weeks. The primary outcome measure was loss of HBeAg at week 78.
During therapy and posttreatment follow-up, patients were monitored monthly by routine physical examination, as well as biochemical and hematological assessments. ALT was assessed locally in accordance with standardized procedures and therefore expressed as times the ULN. Hepatitis flares were defined as a 3-fold increase in serum ALT compared with baseline level. Serum HBV DNA level was measured monthly using an in-house developed Taqman PCR assay (lower limit of detection 373 copies/ml) based on the Eurohep standard. HBeAg and HBsAg (AxSYM, Abbott, Abbott Park, IL) were assessed in weeks 0, 32, and 52 and week 78 (end of follow-up). HBV genotype and YMDD mutation analyses were performed by INNO-LiPA assay (Innogenetics, Gent, Belgium). Because both HBeAg positivity and HBV DNA above 10,000 copies/ml have been described as detrimental, particularly in patients with advanced fibrosis, we defined virologic response in this study as HBeAg seroconversion (loss of HBeAg and appearance of anti-HBe) in combination with serum HBV DNA < 10,000 copies/ml. A baseline biopsy sample was available for 239 patients; biopsy at the end of treatment was optional. Paired biopsies were taken in 110 patients. Histological scoring was performed according to the modified histological activity index (HAI) by an experienced pathologist who was unaware of the chronological order of biopsies, treatment allocation, and outcome measures. Advanced fibrosis was defined as fibrosis score of 4-6 (HAI). Improvement of liver histology was defined as a reduction of at least 2 points in necroinflammatory score (range 0-18) or 1 point in fibrosis score (range 0-6).
Statistical analysis was performed using the SPSS 14.0 program (SPSS Inc., Chicago, IL). The chi-square, Fisher exact test, and Mann-Whitney U test were used where appropriate. The Kaplan-Meier method was used to compare incidence of dose reduction and discontinuation of therapy. The relation between characteristics at baseline and during therapy and relapse and posttreatment response was examined by logistic regression analyses. Univariate analysis was used to assess the importance of prognostic factors. To investigate the independence of these factors, multivariate logistic regression analysis was performed with all characteristics as variables, with a P value < 0.2 in univariate analysis. A P value <0.05 was considered statistically significant.
Of the 239 patients in this study, 70 had advanced fibrosis (29%) and 24 had cirrhosis (10%). Baseline demographics are shown in Table 1. Five of 27 patients without a baseline liver biopsy sample had thrombocytopenia (platelet count range: 133-144 X 109/l), which could indicate cirrhosis. None of the 27 patients had signs of decompensated liver disease; bilirubin was slightly elevated in 2 patients (17.1 and 19.5 umol/l). Patients with advanced fibrosis were older than those without and more often had acquired their infection via sexual or parenteral transmission. Baseline platelet count was significantly lower and necroinflammatory score higher in patients with advanced fibrosis. Genotypes A (32%) and D (40%) were the most prevalent genotypes in our mainly Caucasian patient population. Distribution of genotypes was significantly different between the patient groups. More patients with genotype A had advanced fibrosis (49%) than did patients with genotypes B (19%), C (28%), or D (17%) - P = 0.01, P = 0.03, and P < 0.001, respectively.
Response to Treatment.
At the end of follow-up, all treatment outcomes were observed equally in patients treated with PEG-IFN-a-2b alone and patients treated with PEG-IFN-a-2b in combination therapy with lamivudine. Treatment allocation did not influence response rates in patients with, nor in those without advanced fibrosis. The different response rates of patients with and without advanced fibrosis are given in Table 2. Virologic response at week 78, defined as HBeAg seroconversion and HBV DNA < 10,000 copies/ml, occurred significantly more often in patients with advanced fibrosis than in those without (P = 0.02), as well as decreased fibrosis score (P < 0.001). Although not statistically significant, the rate of HBsAg seroconversion was twice as high in patients with advanced fibrosis compared to those without. Similar trends were observed for improvement of necroinflammation and HBV DNA < 400 copies/ml. Response rates were comparable between patients who had no liver biopsy taken at baseline and those with available baseline liver histology (18% versus 16% for virologic response at week 78; P = 0.76).
Response rates in the subgroup of patients with cirrhosis were similar to those in the group of patients with advanced fibrosis. Thirty-five percent of patients with cirrhosis showed virologic response at week 78 compared with 14% of patients without cirrhosis (P = 0.02). For HBsAg seroconversion, these rates were 13% and 5%, respectively (P = 0.13).
HBV genotype was found to influence outcome at the end of follow-up. The rate of virologic response was higher in patients with genotype A (30%) and B (26%) than in patients with genotype C (3%) and D (6%) infection (A versus C, P = 0.002; A versus D, P < 0.001; B versus C, P = 0.02; B versus D, P = 0.02). Among patients with advanced fibrosis a similar trend was observed for genotype A and D infection (33% versus 8%, P = 0.14). Among patients with genotype A infection, virologic response occurred equally in those with and without advanced fibrosis (33% versus 27%; P = 0.54).
Logistic Regression Analysis of Factors Predicting Response at Week 78.
For the prediction of virologic response at the end of follow-up, the following variables were included in univariate logistic regression analysis: age, sex, weight, race, route of transmission, baseline serum ALT, baseline HBV DNA level, HBV genotype, liver histology, previous interferon or lamivudine therapy, and treatment allocation. Increasing age, sexual or parenteral transmission, presence of advanced fibrosis or cirrhosis, and high HAI score were found to be associated with increased likelihood of virologic response (Table 3). In multivariate analysis, HBV genotype and high necroinflammatory score (RR 1.31, 95% CI 1.05-1.65, P = 0.02) independently predicted virologic response at week 78. HBV genotype was found to be the strongest predictor of virologic response. Genotype A infection was associated with higher response rate than was genotype C infection (RR 11.30, 95% CI 1.38-92.57, P = 0.02) and genotype D infection (RR 4.28, 95% CI 1.39-13.21, P = 0.01). Genotype B infection also resulted in a higher likelihood of response compared to genotype C infection (RR 12.13, 95% CI 1.24-118.30, P = 0.03) and genotype D infection (RR 4.59, 95% CI 1.14-18.43, P = 0.03). In multivariate analysis, the presence of advanced fibrosis did not alter the likelihood of achieving virologic response (RR 0.98, 95% CI 0.17-5.23, P = 0.98).
Most adverse events occurred equally in patients with and without advanced fibrosis (Table 4). Fatigue, anorexia, and thrombocytopenia were observed more frequently in patients with advanced fibrosis than in those without (P < 0.01). These adverse events also occurred more frequently in the subgroup of patients with cirrhosis than in the subgroup of patients without cirrhosis, as well as hepatitis flares (33% versus 12%, P = 0.008) and dizziness (21% versus 7%, P = 0.05). The higher rate of thrombocytopenia in patients with advanced fibrosis did not result in an increase in significant bleeding problems. Nevertheless, epistaxis tended to occur more often in patients with advanced fibrosis (P = 0.08). Necessity for PEG-IFN dose reduction was comparable for patients with advanced fibrosis (33%) and without advanced fibrosis (34%), as well as premature discontinuation of therapy (11% and 8%, respectively; Fig. 1). The reason for dose reduction of PEG-IFN in patients with advanced fibrosis was more often thrombocytopenia than in patients without advanced fibrosis (32% versus 5%, P = 0.01). For both patients with and without advanced fibrosis, most dose reductions took place during the first 16 weeks of therapy (79% and 66%, P = 0.25). Serious adverse events occurred equally in patients with or without advanced fibrosis (4% and 5%, respectively, P = 1.00). In patients with cirrhosis, dose reduction was necessary more often than in patients without cirrhosis (63% versus 30%, P = 0.002), as was premature discontinuation of therapy (21% versus 8%, P = 0.05). Timing of dose adjustments or discontinuation of therapy did not differ between the patient groups.
In this study of the use of PEG-IFN therapy, virologic response, defined as HBeAg seroconversion and HBV DNA <10,000 copies/ml at the end of follow-up, occurred significantly more often in patients with advanced fibrosis than in those without. The rate of HBsAg seroconversion was twice as high in patients with advanced fibrosis, although this difference was not statistically significant.
Because IFN-based therapy is often presumed to be less effective in patients with advanced fibrosis, PEG-IFN is frequently withheld from these patients. In a retrospective study of standard IFN in 200 HBeAg-positive patients, response rates were reported to be significantly lower in patients with advanced fibrosis (Metavir fibrosis score 3-4, 17.5%) than in patients with minimal or no fibrosis (Metavir fibrosis score 0-1, 35.5%; P = 0.02). Lower stage of fibrosis and higher grade of necroinflammation were identified as independent predictors of response to standard IFN in this study. Our results clearly contradict these findings. The higher response rates of patients with advanced fibrosis in our study might have been a result of the antiviral drug used (PEG-IFN versus standard IFN), the longer duration of therapy (1 year versus 4 months), or the predominant HBV genotypes.
Multivariate analysis showed no relation between presence of advanced fibrosis and response to PEG-IFN-a-2b. HBV genotype was identified as the strongest independent predictor of virologic response at the end of follow-up. Previous studies also have shown a higher likelihood of response to (PEG-)IFN therapy of patients harboring genotype A. The higher response rates in patients with advanced fibrosis compared with those without thus appears to also be related to the higher prevalence of genotype A in these patients. Because multivariate analysis showed no influence of advanced fibrosis on response to PEG-IFN therapy, other factors appear to be of greater influence on response rate in these patients. Nevertheless, our finding that advanced fibrosis does not compromise the efficacy of PEG-IFN is important. Furthermore, our findings emphasize the importance of testing for HBV genotype prior to the initiation of antiviral therapy.
Nucleos(t)ide analogs are often considered preferential over PEG-IFN in patients with advanced fibrosis as prolonged treatment with these drugs may reduce progression to decompensated liver disease and development of hepatocellular carcinoma in patients with advanced fibrosis. However, improved survival and decreased incidence of HCC have also been observed in patients with and without preexisting cirrhosis who responded to IFN therapy.
Concerns are often raised about the safety of PEG-IFN in patients with advanced fibrosis. We found PEG-IFN--2b to be safe for these patients; most adverse events were observed equally among patients with and without advanced fibrosis. As expected with lower baseline platelet level, patients with advanced fibrosis had thrombocytopenia more frequently. Thrombocytopenia also more frequently led to dose reduction in these patients, whereas the frequency and timing of both dose reduction and premature discontinuation of therapy were comparable between patient groups. In the subgroup of patients with cirrhosis, dose adjustment and premature treatment discontinuation were more often necessary, but despite less drug exposure, a lower response rate was not observed. Only 1 patient with advanced fibrosis presented with temporarily elevated bilirubin, but other signs of compromised liver function were not encountered.
Another reason why PEG-IFN is not often given to patients with advanced fibrosis is the concern that it might precipitate hepatitis flares and subsequent hepatic decompensation. We found a trend toward higher incidence of hepatitis flares in patients with advanced fibrosis, and flares occurred more frequently in patients with cirrhosis than in those without. However, the flares in our patients, who had well compensated liver disease at the start of therapy, did not lead to any clinical problems. Previous studies on IFN confirm the absence of complications in patients with compensated cirrhosis. Less than 1% of HBeAg-positive patients developed hepatic decompensation, whereas up to 60% of patients included in these trials had cirrhosis. In contrast, IFN therapy leads to significant side effects due to bacterial infection and exacerbation of liver disease in patients with Child's class B or C cirrhosis. In a large European study, 9 cases of fatal hepatic decompensation were identified among 2,490 IFN-treated patients. All these patients had documented cirrhosis, and 5 had signs of hepatic decompensation prior to the start of IFN therapy. These observations clearly show that IFN-based therapy should not be given to hepatitis B patients with signs of hepatic decompensation. The setting of a clinical trial and exclusion of patients with signs of advanced liver disease or portal hypertension could have led to our finding that serious adverse events were observed equally in patients with and without cirrhosis. In our experience, routine monthly follow-up and anticipation of adverse events occurring is sufficient for patients with advanced fibrosis and compensated liver disease.
In conclusion, our findings show that PEG-IFN-a-2b is at least as effective in patients with advanced fibrosis as in those without. Although patients with cirrhosis more often require dose reduction or early discontinuation of therapy, treatment is safe and outcome is similar to patients without cirrhosis. Therefore, when aiming for sustained off-treatment response, patients with advanced fibrosis and well-compensated liver disease, in particular those with genotype A or B, should be offered the possibility of a finite course of PEG-IFN, rather than initiation of possibly indefinite nucleos(t)ide analog therapy.