New scoring system for predicting relapse after lamivudine-induced hepatitis B e-antigen loss in chronic hepatitis B patients

Aim:  In endemic areas of hepatitis B virus (HBV) infection, lamivudine-induced hepatitis B e-antigen (HBeAg) loss is not durable. We developed a scoring system to predict the relapse after lamivudine-induced HBeAg loss.
Methods:  Of 93 HBeAg-positive chronic hepatitis B patients receiving lamivudine therapy, we studied the 30 patients who achieved HBeAg loss. Qualitative polymerase chain reaction assays for HBV DNA were performed after HBeAg loss; if results were negative on two consecutive occasions, lamivudine treatment was stopped. Serum HBV DNA levels at the time of HBeAg loss were determined using stored serum by the Cobas Amplicor HBV Monitor Kit.
Results:  Cumulative relapse rates were 44%, 55.8% and 59.8%, after 1, 2 and 5 years, respectively. Univariate analysis showed that the risk factors for relapse were age (≥ 30 years), HBV DNA level (≥ 60 IU/mL) at the time of HBeAg loss and liver cirrhosis. Independent risk factors for relapse were age (odds ratio OR], 4.3; 95% confidence interval CI], 1.2–17.6) and serum HBV DNA level at the time of HBeAg loss (OR, 5.1; 95% CI, 1.3–21.6). We developed a scoring system based on these independent risk factors. No relapse was observed for patients with a risk score of 0. For patients with a risk score of 1, the relapse rates were 41%, 49% and 56% at 1, 2 and 5 years, respectively. All patients with a risk score of 2 relapsed ( P  = 0.01).
Conclusion:  Our new scoring system may be useful for predicting relapses in patients with lamivudine-induced HBeAg loss. Treatment strategies should be individualized according to risk score.