LAMIVUDINE FOR CHRONIC HEPATITIS B AND ADVANCED LIVER DISEASE histopathologist who was blinded to the treatment Bayer Diagnostics, with a lower limit of detection assignments of 0.7 mEq per milliliter). Results were unavail- Patients were excluded if they had any of the fol- able to the investigators until after the completion lowing: evidence of hepatocellular carcinoma(sus- of the study, and serum HBV DNA assays were not picious foci on hepatic ultrasonography at screen- permitted at the investigators'sites during double ing or a rising serum level of alpha-fetoprotein), blind therapy but were allowed after confirmed a serum alanine aminotransferase level more than HBeAg seroconversion or during open -label lamin 10 times the upper limit of normal, any evidence of udine therapy. Samples collected at baseline, at an hepatic decompensation(as defined by the study nual visits, and at the completion of treatment protocol), autoimmune hepatitis, coinfection with were also analyzed for the presence of ymDd mu- hepatitis C or D virus or human immunodeficiency tations by polymerase-chain-reaction assay and virus, other serious concurrent illness(e.g, alco- restriction-fragment-length polymorphism assay holism, uncontrolled diabetes, or cancer), pancre- Samples collected at all scheduled visits from pa- atic amylase or lipase levels more than two times tients with clinical end points were also tested for the upper limit of normal, an elevated serum creat- YMDD mutations inine level, a hemoglobin level of less than 8 g per deciliter, a white-cell count below 1500 per cubic END POINTS millimeter, a platelet count 50,000 percubic mil- The primary end point was time to disease pro- limeter or less, treatment with immunomodulate- gression, as defined by the first occurrence of an ry or chronic antiviral therapy within the 6 months of the following: an increase of at least 2 points in before screening, treatment with any investigation- the Child-Pugh score(an assessment of the severi- al drug within the 30 days before the study began, ty of liver disease [range, 5 to 15, where 5 indicates or any previous treatment with lamivudine. Wom- good liver function and 15 indicates poor liver func- en who were pregnant were also excluded tion] calculated on the basis of the serum bilirubin and albumin levels, the prothrombin time, and the ASSESSMENTS presence and degree of ascites or encephalopathy), Patients were assessed at baseline, at the end of spontaneous bacterial peritonitis with proven sep- months 1 and 3, and at every three months thereat- sis, renal insufficiency, bleeding gastric or esoph ter for clinical evidence of hepatic decompensation ageal varices, the development of hepatocellular or other complications. They were also questioned carcinoma, ordeath related to liver disease Patients about adverse events, concurrent medications, and with a first clinical end pointwere followed for sub- study drug accountability; blood was taken for he- sequent end points. Any increase in the Child-Pugh matology and biochemistry profiles; serum sam- score due solely to laboratory parameters was con- ples were tested for HBeAg, hepatitis B e antibody, firmed on two consecutive visits at least one month and alpha-fetoprotein; and the prothrombin time apart. For patients with albumin levels below 35 was measured. At baseline and every six months per liter or bilirubin levels greater than 34.2 umol thereafter, serum was assayed for HBsAg and hep- per liter(2 mg per deciliter)at baseline, confirma- atitis B surface antibody, and liver ultrasonogra- tory tests were conducted one week after the first phy was performed. HBeAg seroconversion was test. Renal insufficiency was defined as a decrease considered confirmed if two consecutive samples in creatinine clearance to 50 ml per minute(0.8 ml taken at least a month apart were positive for hepa- per second)or less that was confirmed two times, titis B e antibody and negative for HBeAg Hepatic at least one week apart. Hepatocellular carcinoma ultrasonography and liver biopsy or fine-needle was diagnosed on the basis of results of ultraso- aspiration were performed as clinically indicated nography, selective arteriography, imaging ofhe- to investigate or confirm a diagnosis of hepato- patic tumors during the vascular phase, serum cellular carcinoma levels of alpha-fetoprotein, or by cytologic or his erum samples were collected at baseline and tologic evaluation. The evidence for each end point at months 1, 12, 24, 36, 48, and 60 and analyzed was reviewed and confirmed by a blinded clinical for HBV DNA levels at a central laboratory. HBV end-points committee composed of three interna- dNA was determined by a branched-chain hybridi- tionally recognized hepatologists zation assay (Versant HBV DNA Quantitative Assay, NEnglJMed351715www.nejm.orgocTober7,2004 1523 第14页n engl j med 351;15 www.nejm.org october 7, 2004 lamivudine for chronic hepatitis b and advanced liver disease 1523 histopathologist who was blinded to the treatment assignments. Patients were excluded if they had any of the fol￾lowing: evidence of hepatocellular carcinoma (sus￾picious foci on hepatic ultrasonography at screen￾ing or a rising serum level of alpha-fetoprotein), a serum alanine aminotransferase level more than 10 times the upper limit of normal, any evidence of hepatic decompensation (as defined by the study protocol), autoimmune hepatitis, coinfection with hepatitis C or D virus or human immunodeficiency virus, other serious concurrent illness (e.g., alco￾holism, uncontrolled diabetes, or cancer), pancre￾atic amylase or lipase levels more than two times the upper limit of normal, an elevated serum creat￾inine level, a hemoglobin level of less than 8 g per deciliter, a white-cell count below 1500 per cubic millimeter, a platelet count of 50,000 per cubic mil￾limeter or less, treatment with immunomodulato￾ry or chronic antiviral therapy within the 6 months before screening, treatment with any investigation￾al drug within the 30 days before the study began, or any previous treatment with lamivudine. Wom￾en who were pregnant were also excluded. assessments Patients were assessed at baseline, at the end of months 1 and 3, and at every three months thereaf￾ter for clinical evidence of hepatic decompensation or other complications. They were also questioned about adverse events, concurrent medications, and study drug accountability; blood was taken for he￾matology and biochemistry profiles; serum sam￾ples were tested for HBeAg, hepatitis B e antibody, and alpha-fetoprotein; and the prothrombin time was measured. At baseline and every six months thereafter, serum was assayed for HBsAg and hep￾atitis B surface antibody, and liver ultrasonogra￾phy was performed. HBeAg seroconversion was considered confirmed if two consecutive samples taken at least a month apart were positive for hepa￾titis B e antibody and negative for HBeAg. Hepatic ultrasonography and liver biopsy or fine-needle aspiration were performed as clinically indicated to investigate or confirm a diagnosis of hepato￾cellular carcinoma. Serum samples were collected at baseline and at months 1, 12, 24, 36, 48, and 60 and analyzed for HBV DNA levels at a central laboratory. HBV DNA was determined by a branched-chain hybridi￾zation assay (Versant HBV DNA Quantitative Assay, Bayer Diagnostics, with a lower limit of detection of 0.7 mEq per milliliter). Results were unavail￾able to the investigators until after the completion of the study, and serum HBV DNA assays were not permitted at the investigators’ sites during double￾blind therapy but were allowed after confirmed HBeAg seroconversion or during open-label lamiv￾udine therapy. Samples collected at baseline, at an￾nual visits, and at the completion of treatment were also analyzed for the presence of YMDD mu￾tations by polymerase-chain-reaction assay and restriction-fragment–length polymorphism assay. Samples collected at all scheduled visits from pa￾tients with clinical end points were also tested for YMDD mutations. end points The primary end point was time to disease pro￾gression, as defined by the first occurrence of any of the following: an increase of at least 2 points in the Child–Pugh score (an assessment of the severi￾ty of liver disease [range, 5 to 15, where 5 indicates good liver function and 15 indicates poor liver func￾tion] calculated on the basis of the serum bilirubin and albumin levels, the prothrombin time, and the presence and degree of ascites or encephalopathy), spontaneous bacterial peritonitis with proven sep￾sis, renal insufficiency, bleeding gastric or esoph￾ageal varices, the development of hepatocellular carcinoma, or death related to liver disease. Patients with a first clinical end point were followed for sub￾sequent end points. Any increase in the Child–Pugh score due solely to laboratory parameters was con￾firmed on two consecutive visits at least one month apart. For patients with albumin levels below 35 g per liter or bilirubin levels greater than 34.2 µmol per liter (2 mg per deciliter) at baseline, confirma￾tory tests were conducted one week after the first test. Renal insufficiency was defined as a decrease in creatinine clearance to 50 ml per minute (0.8 ml per second) or less that was confirmed two times, at least one week apart. Hepatocellular carcinoma was diagnosed on the basis of results of ultraso￾nography, selective arteriography, imaging of he￾patic tumors during the vascular phase, serum levels of alpha-fetoprotein, or by cytologic or his￾tologic evaluation. The evidence for each end point was reviewed and confirmed by a blinded clinical end-points committee composed of three interna￾tionally recognized hepatologists. 第 14 页