The NEW ENGLAND JOURNAL Of MEDICINE Table 1 Baseline Characteristics of the Patients Lamivudine Group Placebo Grou Characteristic Range 17-74 341(78) 156(73) 75(17) 41(19) 20(5) 18(8) Ishak fibrosis score (%) 127(29) 55(26) (31) V DNA 11.7 21.5 <0.7-109800 HBV DNA 20.7 mEq/ml-no(9)5 345(79) 174(81) 252(58) 124(58) Alpha-fetoprotein -ug/liter 0.7-600 Alanine aminotransferase-U/liter Median 338(78) 171(80) viewing interim analyses. The board was empow All adverse events, regardless of their possible as- ered to recommend termination of the study on the sociation with the disease or study treatment, were basis of safety concerns or as soon as sufficient ev recorded. Adverse events were considered to be se- idence indicated that lamivudine was statistically rious if the investigator determined that they jeop- superior to placebo or that lamivudine did not pro- ardized the patient, were life-threatening, or would vide a significant advantage over placebo esult in hospitalization, disability, or death. STATISTICAL ANALYSES DATA AND SAFETY MONITORING BOARD Sample size was determined on the basis of the pri- The data and safety monitoring board consisted mary analysis of time to disease progression. To es- of three independent hepatologists, who were not timate power, the annual rate of disease progres- members of the end-points committee, and an in- sion was assumed to be 20 percent for the placebo dependent statistician. The board protected the group, 9,29 whereas a reduction in this rate of one ethical interests and safety of the patients by re- third(to 13. 3 percent) for the lamivudine group 524 N englj Med 35715 Www.nejm.org OctobeR 7, 2004 第15页n engl j med 351;15 www.nejm.org october 7, 2004 The new england journal of medicine 1524 safety All adverse events, regardless of their possible as￾sociation with the disease or study treatment, were recorded. Adverse events were considered to be se￾rious if the investigator determined that they jeop￾ardized the patient, were life-threatening, or would result in hospitalization, disability, or death. data and safety monitoring board The data and safety monitoring board consisted of three independent hepatologists, who were not members of the end-points committee, and an in￾dependent statistician. The board protected the ethical interests and safety of the patients by re￾viewing interim analyses. The board was empow￾ered to recommend termination of the study on the basis of safety concerns or as soon as sufficient ev￾idence indicated that lamivudine was statistically superior to placebo or that lamivudine did not pro￾vide a significant advantage over placebo. statistical analyses Sample size was determined on the basis of the pri￾mary analysis of time to disease progression. To es￾timate power, the annual rate of disease progres￾sion was assumed to be 20 percent for the placebo group,8,9,29 whereas a reduction in this rate of one third (to 13.3 percent) for the lamivudine group Table 1. Baseline Characteristics of the Patients.* Characteristic Lamivudine Group (N=436) Placebo Group (N=215) Male sex — no. (%) 370 (85) 182 (85) Asian — no. (%) 426 (98) 210 (98) Age — yr Median 43 44 Range 17–74 22–71 Child–Pugh score — no. (%)† 5 341 (78) 156 (73) 6 75 (17) 41 (19) ≥7 20 (5) 18 (8) Ishak fibrosis score — no. (%)‡ 4 176 (40) 76 (35)‡ 5 127 (29) 55 (26) 6 133 (31) 84 (39) HBV DNA — mEq/ml Median 11.7 21.5 Range <0.7–109,800 <0.7–4234 HBV DNA ≥0.7 mEq/ml — no. (%)§ 345 (79) 174 (81) Positive for HBeAg — no. (%) 252 (58) 124 (58) Alpha-fetoprotein — µg/liter Median 8.6 9.8 Range 0.7–600 1.2–298 Albumin — g/liter Median 42 41 Range 28–54 27–52 Alanine aminotransferase — U/liter Median 70 68 Range 14–959 7–821 Alanine aminotransferase >1 time the upper limit of normal — no. (%) 338 (78) 171 (80) 第 15 页