Interferon in HBV/HIV Coinfection; Serum Alanine Transaminase Level Is a Good Predictor of Response to Interferon Alfa Therapy for Chronic Hepatitis B in Human Immunodeficiency Virus-Infected Patients

Hepatology, April 2000, p. 1030-1030, Vol. 31, No. 4

To the Editor:

In a recent issue of HEPATOLOGY, Benhamou et al. reported a high incidence of hepatitis B virus (HBV) resistance to lamivudine in human immunodeficiency virus

(HIV)-infected patients, close to 100% at 5 years, underlying the need for alternative therapies. Indeed, although HIV-related immunodepression may reduce liver necroinflammatory lesions and serum alanine transaminase (ALT) level in HBV-infected patients, HBV infection has been shown to be associated with more severe liver fibrosis in HIV-coinfected patients, which may increase mortality. Interferon alfa therapy, which was previously used as the most effective therapy in chronic HBV-infected patients, has been reported to be poorly efficient in HIV-infected patients, with a hepatitis B e (HBe) seroconversion rate close to 0%. However, few cases of persistent interferon-induced HBe seroconversion have been reported in HIV-infected patients, and the factors associated with the response to interferon therapy have never been well documented in HIV-HBV coinfected patients.

From 1987 to 1996, 26 HIV-positive homosexual men received a single 6-month course of interferon alfa therapy, 5 millions-units, 3 times weekly for chronic hepatitis B in H┘pital Beaujon (Clichy, France). They all were infected by a wild-type HBV, had neither HDV nor HCV coinfection, and had never received lamivudine therapy. Their characteristics on starting interferon therapy were as follows (median, range): age: 34 years, 26-55; HBV infection duration: 19 months, 6-113; serum ALT level: 2.1 upper limit of normal (ULN), 1.1-19.5; CD4 cell count: 295/mm3, 13-800; serum HBV-DNA level (Genostics, Abbott, Rungis, France): 241 pg/mL, 5-1220; histological activity index (Knodell scoring system): 2-17; with necroinflammatory lesions (Knodell scoring system) at 7, 1-14. Five patients consumed alcohol (more than 40 g/d). Three had cirrhosis. At the end of therapy, a loss of serum HBV DNA was observed in 7 cases (27%). A complete HBe seroconversion occurred during interferon therapy or during the 6 months after the end of therapy in 4 cases (15%). All of these patients remained anti-HBe antibody positive at 2 years of follow-up and showed histological improvement. When considering all the available characteristics at baseline, the only factor associated with both serum HBV-DNA loss and HBe seroconversion was a high pretherapeutic serum ALT level. Serum ALT level was 7.5 versus 2.9 ULN (Mann-Whitney U test, P = .005) in patients who experienced HBV-DNA loss, and was 10.7 versus 2.9 ULN (P = .004) in patients who experienced HBe seroconversion. No significant difference was assessed for age, previous HBV infection duration, histological activity index, or proportion of alcohol consumers. Although higher, the CD4 cell count was not significantly different in patients who experienced HBV-DNA loss (420 vs. 328/mm3, not significant) or in patients who experienced HBe seroconversion (492 vs. 314/mm3, not significant). High serum ALT levels between 3 ULN and 10 ULN were associated with a significant risk of interferon-induced HBV-DNA loss and HBe seroconversion.

Our results suggest that interferon therapy can be successfully used for chronic hepatitis B in HIV-infected patients with high serum ALT levels, as previously reported in HIV-negative patients. Hence, the follow-up of serum ALT levels in HIV-infected patients with chronic hepatitis B would be useful to select patients that are potentially good responders to interferon therapy and to determine the favorable time for initiating therapy. Such information may be useful for HIV-infected patients who develop HBV resistance to lamivudine. Finally, the efficacy of pegylated-interferon needs also to be evaluated in this population.