Hepatitis C: Somber Views of Natural History and Optimistic Views of Interferon Treatment?
HEPATOLOGY, May 1998, p. 1443-1444, Vol. 27, No. 5
SEE ARTICLE 1435
Hepatitis C is a slow and silent killer, which, after a period of considerable controversy, is now clearly recognized as a cause of excess mortality. The most convincing evidence of serious nature of Hepatitis C virus (HCV) infection is provided by a study from the United Kingdom of a cohort of hemophiliacs in which 25 years of infection was associated with a seventeen-fold increased risk of death from liver disease and a sixfold increased risk of liver cancer compared with those without infection.1 Also, national mortality data from Europe suggest a dramatic increase in primary liver cancer. 2,3 In France between 1979 and 1994, the period concurrent with the French epidemic of Hepatitis C, death caused by primary liver cancer increased fourfold in males and twofold in females. 4,5 This evidence of excess mortality brings a sense of urgency to the assessment of the efficacy of treatment, particularly in patients with HCV-associated cirrhosis. To date, only one randomized study of interferon (IFN) has been published which shows a significant difference in the incidence of hepatocellular carcinoma (HCC) between treated and untreated patient6; however, this study is isolated and is imperfect in design, with concerns being raised regarding the small sample size and the relatively short duration of therapy. Clearly, further studies are needed to assess the effect of interferon on major clinical endpoints (ie, the development of complications of cirrhosis such as HCC and hepatic decompensation) and on development of death.
Two nonrandomized studies which address the possible beneficial role of interferon, one from Japan7 and one from France,8 are published in this issue of HEPATOLOGY. Although very different in design, both emphasize the risk of HCC in patients with cirrhosis, independent of genotype, and both suggest a possible beneficial effect of IFN in reducing the incidence of HCC.
The paper by Kasahara et al.7 describes a multi-center study of patients treated for Hepatitis C, most of whom do not have cirrhosis. Its strengths are as follows: a large sample size (1022 patients treated with IFN); a largely prospective study design; the inclusion of data on genotype and viral load; the exclusion of patients receiving low doses of IFN that is less than a total of 200 MU, which corresponds to the standard regimen of 3 MU TIW for 6 months.
The low prevalence of cirrhosis (less than 3%) present in the study population initially may explain the relatively large amounts of IFN which could be administered (500 MU), as patients with cirrhosis are typically less tolerant of IFN than those without cirrhosis. Moreover, these increased doses may also explain the 30% rate of sustained response, a percentage which is higher than that traditionally observed with standard doses of IFN. There are a number of limitations of this study, however, including the following: the lack of a prospective control group, lack of details concerning the timing of tests such as ultrasonography, -fetoprotein levels, histological assessment, inadequate information concerning alcohol consumption, and lack of detail regarding the type of IFN used. Finally, the period of follow-up at 39 months was relatively short for a non-cirrhotic population. The message for the clinician is that there appears to be a relationship between the response to IFN and the reduced incidence of HCC. The benefit of interferon was observed both in transient and sustained responders but not in non-responders. This study strongly supports the use of 3 MU TIW for 12 months, for a total administered dose of approximately 500 MU.
The benefit of IFN may be secondary to a decrease in progression of liver disease to cirrhosis in responders compared with non-responders. Meta-analyses of randomized trials 9,10 have shown that IFN significantly reduces the grade of inflammatory activity and stage of fibrosis in non-cirrhotic patients, particularly in those who respond to treatment. A recent randomized trial has also demonstrated that an aggressive IFN regimen reduces the incidence of cirrhosis despite a very low sustained virological response rate.11 In the study of Kasahara et al.,7 the progression of fibrosis was not assessed by follow-up biopsy. The correlation between the degree of fibrosis present initially and the development of HCC was almost significant at P = .052. By univariate analysis, HCC developed more frequently in patients with advanced liver fibrosis than in patients with mild fibrosis (P = .0008). Even in those who did not respond to IFN, a treatment benefit could exist, although the lack of an untreated control group makes the determination of any effect difficult. In the non-responders, the incidence of HCC was 3.7% per year, compared with 5% to 7% per year observed in Japanese patients with cirrhosis. In the absence of an adjustment for the incidence of cirrhosis, this difference is impossible to evaluate.
In the second study in this issue of HEPATOLOGY, Serfaty et al. evaluated retrospectively the effects of IFN in a selected population of HCV-positive patients with compensated cirrhosis. The strengths of this study are as follows: 1) criteria for inclusion were clearly defined; 2) treatment was with either interferon alpha 2a or 2b; 3) only 3% of patients were lost to follow-up; 4) an untreated comparison group was included; and 5) most importantly, all complications of liver disease such as variceal hemorrhage, hepatic insufficiency, need for liver transplantation as well as development of HCC were evaluated. Its weaknesses include the following: a lack of randomization as well as omission of several potentially relevant details which may influence rate of development of complications and/or mortality, such as duration of follow-up; use of screening ultrasonography and -fetoprotein levels; quantitation of alcohol consumption; and administration of adjunctive therapy, such as beta-blockers or endoscopic sclerotherapy for portal hypertension. Moreover, the total dose of IFN received and the duration of treatment in non-responders was not provided. The principal message for the clinician from this second study is that, at about 3% per year, the incidence of HCC in patients with Hepatitis C cirrhosis is similar worldwide, an incidence which is higher than previous estimates, and that this risk is independent of genotype responsible for infection.12 Results from this study also suggest that IFN may even benefit patients with HCV cirrhosis who traditionally rarely achieve a sustained virological response rate. These data are consistent with the results of a randomized trial from our group, in which histological improvement was observed in 80% of cirrhotics who received long-term IFN compared with only 24% in patients who were treated for only 6 months.13 Serfaty et al. also point out the need for markers of disease progression, suggesting that serum albumin could be a useful predictor of future hepatic decompensation.
Thierry Poynard, M.D., Ph.D
Pierre Opolon, M.D.
Groupe Hospitalier Pitie Salpêtrière
Abbreviations: HCV, Hepatitis C virus; IFN, interferon; HCC, hepatocellular carcinoma.
From the Service Hepato-Gastroenterologie, Groupe Hospitalier Pitie Salpêtrière, Paris, France
Received January 16, 1998; accepted March 3, 1998.