Patients with hepatocellular carcinoma sometimes have erythrocytosis and high plasma erythropoietin levels. However, previous studies have not revealed direct evidence that the carcinoma cells produce the erythropoietin. To address this question, we carried out light and electron microscopic immunohistochemical studies, using a human erythropoietin antibody to the liver in three male patients with hepatocellular carcinoma and erythrocytosis. alpha-Feto-protein localization was also examined in serial liver sections by light microscopic immunohistochemistry with an antibody to alpha-fetoprotein. All three patients demonstrated high hemoglobin levels (16.7, 17.6 and 18.1 gm/dl) and high plasma erythropoietin levels (227, 266 and 280 mU/ml). In one patient the plasma erythropoietin level in the hepatic vein was significantly higher than that in the hepatic artery. The levels of plasma erythropoietin, as well as such tumor markers for hepatocellular carcinoma as serum alpha-fetoprotein and plasma des-gamma-carboxyprothrombin, were significantly reduced after treatment with an anticancer drug, cisplatin. Light microscopic immunohistochemistry showed that erythropoietin was definitely present in the cytoplasm of the hepatocellular carcinoma cells, but not in normal hepatocytes around the carcinoma lesion or in other nonparenchymal cells such as vascular endothelial cells and Kupffer cells. In electron microscopic immunohistochemistry, reaction products for erythropoietin were revealed in the cisternae of the endoplasmic reticulum in the carcinoma cells, suggesting the production of erythropoietin by these cells. Light microscopic immunohistochemistry showed that alpha-fetoprotein was localized in the hepatocellular carcinoma cells that were erythropoietin positive in the serial sections. These findings indicated that hepatocellular carcinoma cells produced erythropoietin as well as alpha-fetoprotein in these cases, leading to the complication of erythrocytosis.

No abstract

It has been reported that spontaneous or interferon (IFN)-induced hepatitis B e (HBe) seroconversion has usually been associated with the development of a stop codon in the precore region. However, the difference between lamivudine-induced seroconversion and spontaneous or IFN-induced seroconversion is not known. The aim of this study was to investigate the correlation between the evolution of the precore and core promoter mutations and lamivudine-induced seroconversion. Forty-five patients with chronic hepatitis B virus (HBV) infection who were treated with lamivudine for more than 1 year were enrolled. The nucleotide sequence of the precore and core promoter region was determined before and after treatment with lamivudine for 1 year. Among 29 patients who were hepatitis B e antigen (HBeAg)-positive before treatment, 12 (41.3%) lost HBeAg during the course of treatment for 1 year. Of these, eight patients (66.7%) still had precore wild type HBV after 1 year. After 1 year, reversion to precore wild type HBV was detected in 11 (64.7%) of 17 patients who had precore mutant HBV before treatment. Twelve (70.6%) of 17 patients who were persistently HBeAg-positive had precore wild type HBV before and after treatment for 1 year. Despite the loss of HBeAg, two thirds of the patients still had precore wild type HBV after the 1-year treatment. It is suggested that lamivudine-induced seroconversion differs from spontaneous or IFN-induced seroconversion in the change of nucleotides in the precore region. The reversion in the precore region may be caused by the difference of drug-susceptibility to lamivudine. The antiviral effect of lamivudine may be more effective in the precore mutant HBV than in the precore wild type HBV.

There have been reports of relapse after cessation of lamivudine monotherapy for hepatitis B virus (HBV) infection. The aim of this study was to examine factors that predict posttreatment relapse. Comparison 22 patients who experienced relapse with 11 who did not after cessation of therapy showed that predictive factors for nonrelapse were hepatitis B e antigen seroconversion and duration of undetectable HBV DNA load (<0.7 log IU/mL), as determined by HBV real-time detection direct testing. However, 7 of 12 patients with seroconversion experienced relapse after cessation of therapy. Multivariate analysis revealed that the duration of an undetectable HBV DNA load was the only independent predictive factor for nonrelapse (odds ratio, 0.50; 95% confidence interval, 0.27-0.9). More-prolonged lamivudine therapy is required after seroconversion, and persistent duration of an HBV DNA level of <0.7 log IU/mL for >6 months can more accurately aid in the decision of when to stop lamivudine therapy.

The extended treatment significantly increased the SVR rate in patients who were HCV RNA negative at 16-24 weeks.