Chronic hepatitis B virus (HBV) infection affects

Transcription

1 GASTROENTEROLOGY 2009;136: Year GLOBE Trial Results: Telbivudine Is Superior to Lamivudine in Patients With Chronic Hepatitis B YUN FAN LIAW,* EDWARD GANE, NANCY LEUNG, STEFAN ZEUZEM, YUMING WANG, CHING LUNG LAI, # E. JENNY HEATHCOTE,** MICHAEL MANNS, NATALIE BZOWEJ, JUNQI NIU, STEVEN HUY HAN, SEONG GYU HWANG, ## YILMAZ CAKALOGLU,*** MYRON J. TONG, GEORGE PAPATHEODORIDIS, YAGANG CHEN, NATHANIEL A. BROWN, EFSEVIA ALBANIS, ### KARIN GALIL, and NIKOLAI V. NAOUMOV ### ; The GLOBE Study Group *Liver Research Unit, Chang Gung Memorial Hospital, Chang Gung University College of Medicine, Taipei, Taiwan; Department of Gastroenterology, Middlemore Hospital, Auckland, New Zealand; Department of Medicine, Alice Ho Miu Ling Nethersole Hospital, Hong Kong, China; Department of Medicine 1, J.W. Goethe Univeristy Hospital, Frankfurt, Germany; Xi Nan Hospital, Third Military Medical University, Chongqing, China; # University Department of Medicine, Queen Mary Hospital, Hong Kong, China; **Department of Medicine, Division of Gastroenterology, University of Toronto, Toronto, Ontario, Canada; Department of Gastroenterology, Hepatology and Endocrinology, Hannover Medical School, Hannover, Germany; Department of Liver Transplantation, California Pacific Medical Center, San Francisco, California; Department of Hepatology, First Hospital, University of Jilin, Guangzhou, China; Pfleger Liver Institute, UCLA School of Medicine, Los Angeles, California; ## Department of Internal Medicine, Pochon CHA University, Bundang Cha Hospital, Kyunggi-Do, Republic of Korea; ***Istanbul University, Istanbul, Turkey; Phleger Liver Institute, Division of Digestive Diseases, David Geffen School of Medicine, University of California, Los Angeles, California; 2nd Department of Internal Medicine, Athens University Medical School, Hippokration Hospital, Athens, Greece; First Affiliated Hospital, Zhejiang University, Hangzhou, China; Idenix Pharmaceuticals, Cambridge, Massachusetts; and ### Novartis AG, Basel, Switzerland See editorial on page 389. Background & Aims: The GLOBE trial has compared the efficacy and safety of telbivudine versus lamivudine treatment over 2 years in patients with chronic hepatitis B. Methods: Hepatitis B e antigen (HBeAg)-positive (n 921) and HBeAg-negative (n 446) patients received telbivudine or lamivudine once daily for 104 weeks. The primary outcome, assessed in the intent-to-treat population, was therapeutic response (hepatitis B virus DNA <5 log 10 copies/ml and either HBeAg loss or normalization of alanine aminotransferase [ALT] level). Results: The therapeutic response to telbivudine was superior to that of lamivudine in HBeAg-positive (63% vs 48%; P <.001) and HBeAg-negative (78% vs 66%; P.007) patients. HBeAg-positive patients given telbivudine also had better outcomes compared with lamivudine in terms of nondetectable viremia (<300 copies/ml) at 55.6% versus 38.5% (P <.001), HBeAg loss at 35.2% versus 29.2% (P.056), and viral resistance at 25.1% versus 39.5% (P <.001). Hepatitis B e antigen seroconversion was 29.6% versus 24.7% (P.095) in all patients and 36% versus 27% (P.022) in patients with baseline ALT level >2 times normal. Telbivudine-treated HBeAg-negative patients showed higher rates of nondetectable viremia compared with lamivudine at 82.0% versus 56.7% (P <.001) and less resistance at 10.8% versus 25.9% (P <.001). Adverse events occurred with similar frequency, whereas grade 3/4 increases in creatine kinase levels were more common in patients given telbivudine (12.9% vs 4.1%, P <.001). Multivariate logistic regression analyses identified telbivudine treatment, among other variables, as an independent predictor of better week 104 outcomes. Conclusions: Telbivudine is superior to lamivudine in treating patients with chronic hepatitis B over a 2-year period. Chronic hepatitis B virus (HBV) infection affects million individuals worldwide and is particularly prevalent in the Asian-Pacific region. 1 Persistent HBV replication increases the risk of cirrhosis, hepatocellular carcinoma, and liver failure. 2,3 Serum HBV DNA is a marker of viral replication, and several natural history studies have shown that high serum HBV DNA levels are associated with an increased risk of disease progression. 4 7 Consistent with these findings, current disease management guidelines indicate that effective and prolonged viral suppression is a primary aim of antiviral treatment of patients with chronic hepatitis B. 8,9 In addition to lamivudine, there are currently 4 oral nucleos(t)ide analogues available for treatment of chronic hepatitis B, including adefovir dipivoxil, entecavir, telbivudine, and recently approved tenofovir, which have shown greater antiviral activity compared with placebo or lamivudine in patients with chronic hepatitis B However, clinical experience with anti-hbv nucleos(t)ides has shown that therapy extended beyond 1 year is necessary for most Abbreviations used in this paper: CK, creatine kinase; PCR, polymerase chain reaction; ULN, upper limit of normal by the AGA Institute /09/$36.00 doi: /j.gastro

2 February 2009 TELBIVUDINE VS LAMIVUDINE FOR HEPATITIS B 487 patients. 9 Emerging evidence suggests that the patients most likely to achieve optimal therapeutic outcomes may be predicted by their initial virologic response during the first 6 months of treatment After 1 year of treatment in the phase 3 GLOBE trial, telbivudine demonstrated significantly greater antiviral activity and less resistance compared with lamivudine in both hepatitis B e antigen (HBeAg)-positive and HBeAgnegative patients, a result that formed the basis of the global registrations for telbivudine. 16 The GLOBE trial was designed as a 2-year intent-to-treat comparison of telbivudine and lamivudine, and here we report the final 2-year safety and efficacy results from this trial. We also explore relationships between baseline HBV DNA and alanine aminotransferase (ALT) levels and early virologic response (HBV DNA levels at treatment week 24) with the outcomes at 2 years. Patients and Methods Patients and Study Design Between March 2003 and April 2004, a total of 1367 adults aged years with HBeAg-positive or HBeAg-negative chronic hepatitis B and compensated liver disease were enrolled in this prospective, randomized, double-blind phase 3 trial, as reported previously. 16 Key enrollment criteria included serum ALT level times the upper limit of normal (ULN), serum HBV DNA level 6 log 10 copies/ml, and no prior treatment with anti-hbv nucleosides or nucleotides. Eligible subjects were randomized using a computer-generated code via central telephone (1:1) to receive telbivudine (600 mg) or lamivudine (100 mg), each with matching placebo, once daily as oral tablets for 2 years. Treatment assignments were stratified by HBeAg status (positive or negative) and serum ALT level ( or 2.5 times the ULN). Efficacy and Safety Assessments Therapeutic response, the primary efficacy end point, was defined as reduction of serum HBV DNA levels to 5 log 10 copies/ml, as recommended in treatment guidelines available when this study was designed, coupled with either normalization of ALT level or loss of serum HBeAg. 2,3 Secondary efficacy measures included serum HBV DNA changes from baseline, proportion of patients with HBV DNA nondetectable by polymerase chain reaction (PCR) assay ( 300 copies/ml), HBeAg loss and seroconversion, normalization of serum ALT level, and hepatitis B surface antigen (HBsAg) loss and seroconversion. Liver histology was evaluated at enrollment and week 52 but not at week 104. An independent data safety monitoring board and an external data analysis center monitored safety. Adverse events and laboratory abnormalities were monitored throughout the 104-week treatment phase and graded for severity according to criteria adapted from the Division of Acquired Immunodeficiency Syndrome, National Institute of Allergy and Infectious Diseases. 24 Grade 3 aminotransferase elevations were defined as 3 10 times baseline; grade 4 elevations were defined as 10 times baseline and/or evidence of hepatic failure. Grades 3 4 creatine kinase (CK) elevations were defined as levels 7 times the ULN. ALT flares were defined as elevations 2 times baseline or 10 times the ULN, as per American Association for the Study of Liver Diseases practice guidelines. 9 HBV Genotyping, Viral Breakthrough, and Resistance Viral breakthrough was defined as a persistent (2 consecutive determinations) on-treatment increase in serum HBV DNA level 1 log 10 copies/ml above the nadir level. 16 Resistance was defined as viral breakthrough with the emergence of a treatment-associated resistance mutation, identified by sequencing of the HBV polymerase gene in the amplified HBV DNA from sera at baseline and week 104 as described previously. 16,25 In addition, the entire reverse transcriptase domain of HBV polymerase gene was sequenced from sera of all patients with residual serum HBV DNA level 3 log 10 copies/ml at week 104. All HBV DNA sequencing and genotyping were performed at an independent laboratory (Delft Diagnostic Laboratory, Voorburg, The Netherlands). Primary treatment failure was defined as completion of at least 24 weeks of treatment without 2 consecutive measurements of serum HBV DNA level 5 log 10 copies/ml. Durability of HBeAg Loss HBeAg-positive patients were eligible for efficacyrelated treatment discontinuation if they received at least 52 weeks of therapy, were HBeAg negative for at least 24 weeks, and had a serum HBV DNA level 5 log 10 copies/ ml. Treatment discontinuation was not compulsory for patients who met these criteria but was at the discretion of the investigators due to varying local practices. Statistical Analysis The intent-to-treat population comprised all randomized patients who received at least one dose of study medication. For categorical efficacy end points, missing values were considered failures. The primary study analysis occurred at week 52, with assessments at week 104 considered important for assessing the longer-term comparative safety and efficacy of telbivudine versus lamivudine. Because the patient population had compensated liver disease, with expected low rates of hepatic decompensation, no interim analyses were planned in the protocol. Study treatments were compared for noninferiority; if prospectively defined noninferiority criteria were met for the primary efficacy end point (therapeutic response), superiority assessments were undertaken. A weighted

3 488 LIAW ET AL GASTROENTEROLOGY Vol. 136, No. 2 Cochran Mantel Haenszel method, adjusting for randomization strata, was used to assess comparative therapeutic response rates. 26 Patients were stratified by HBeAg status (positive or negative) at randomization to accommodate a possible interaction between treatment effect and HBeAg status for therapeutic responses. Such an interaction was evident at week 52 at the 0.15 significance level; results are therefore presented separately for the HBeAg-positive and HBeAg-negative patient populations. Confidence intervals (95.68% for the primary end point, 95% for others) were calculated to evaluate statistical noninferiority for telbivudine versus lamivudine. For continuous variables, analysis of variance was performed with each stratified group as a factor. Reported P values are 2 sided and not adjusted for multiple testing. A P value of.0432 for the primary end point or a P value of.05 for others indicates a statistically significant treatment difference. Results A total of 1370 patients were randomized to treatment from 112 academic centers in 20 countries on 4 continents. The intent-to-treat population comprised 921 HBeAg-positive and 446 HBeAg-negative patients (Figure 1). Treatment groups were well matched at baseline with respect to demographic and disease parameters (Table 1). In the trial, 56 (8%) and 88 (13%) patients receiving telbivudine and lamivudine, respectively, withdrew before week 104 (Figure 1). No patient in the telbivudine arm discontinued due to disease progression versus 2 patients in the lamivudine group. Efficacy Outcomes Therapeutic and HBV DNA response. In HBeAgpositive and HBeAg-negative patients at week 104, therapeutic response was achieved by significantly more recipients of telbivudine than lamivudine (Table 2). Reductions in serum HBV DNA level from baseline to week 104 were significantly greater with telbivudine compared with lamivudine in HBeAg-positive and HBeAgnegative patients (Table 2). At week 104, serum HBV DNA was nondetectable in significantly more patients treated with telbivudine versus Figure 1. Consort flowchart for all HBsAg patients enrolled into the GLOBE trial. IVRS, interactive voice response system; E, HBeAg positive; E-ve, HBeAg negative; ITT, intent to treat; LOCF, last observation carried forward.

4 February 2009 TELBIVUDINE VS LAMIVUDINE FOR HEPATITIS B 489 Table 1. Baseline Demographics and Clinical Characteristics (Intent-to-Treat Population) HBeAg-positive patients HBeAg-negative patients Characteristic Telbivudine (n 458) Lamivudine (n 463) Telbivudine (n 222) Lamivudine (n 224) Mean age, range (y) 32 (16 63) 33 (16 67) 43 (17 68) 43 (18 68) Male, n (%) 333 (73) 351 (76) 174 (78) 177 (79) Mean weight, range (kg) 66 (38 126) 68 (38 150) a 72 (42 123) 71 (45 148) Race, n (%) Chinese 265 (58) 265 (57) 116 (52) 102 (46) Non-Chinese 115 (25) 106 (23) 29 (13) 42 (19) Asian 52 (11) 55 (12) 46 (21) 56 (25) White 4 (1) 7 (2) 3 (1) 3 (1) Black 2 ( 1) 4 (1) 2 (1) 4 (2) Latino 8 (2) 7 (2) 6 (3) 4 (2) Middle 12 (3) 19 (4) 20 (9) 13 (6) Eastern/Indian Other HBV genotype, n (%) A 24 (5) 31 (7) 12 (5) 14 (6) B 129 (28) 113 (24) 59 (27) 59 (26) C 259 (57) 258 (56) 89 (40) 86 (38) D 42 (9) 54 (12) 57 (26) 64 (29) Other or unknown 3 (1) 7 (2) 5 (2) 1 (0.4) Serum ALT level (IU/L) b Mean SE Median (range) ( ) ( ) 99.0 (31 569) 98.5 (12 982) Serum ALT level 2 times the ULN (%) 295 (64) 293 (63) 130 (59) 125 (56) Serum HBV DNA level (log 10 copies/ml) Mean SE Median (range) 9.6 ( ) 9.6 ( ) 7.2 ( ) 7.1 ( ) a Data were missing for 1 patient. b The ULN for serum ALT level was 48 IU/L for men and 37 IU/L for women. lamivudine in HBeAg-positive patients and HBeAg-negative patients (Table 2). The mean time required to achieve nondetectable HBV DNA was significantly shorter with telbivudine versus lamivudine in HBeAg-positive patients (34 vs 39 weeks; P.001) and also in HBeAg-negative patients (20 vs 26 weeks; P.001). Normalization of ALT levels. The rates of serum ALT normalization at week 104 were 70% and 62% among HBeAg-positive patients treated with telbivudine and lamivudine, respectively (P.05; Table 2). In HBeAg-negative patients, normalization of ALT level by week 104 was achieved by 78% and 70% of telbivudine and lamivudine recipients, respectively. Among HBeAgpositive patients with nondetectable serum HBV DNA at week 24, 82% and 79% achieved normalization of ALT level by week 104 with telbivudine and lamivudine, respectively. HBeAg and HBsAg responses. In all HBeAg-positive patients, a larger proportion of telbivudine recipients experienced HBeAg loss compared with lamivudine (P.056; Table 2). The rates of HBeAg loss and seroconversion were proportionally greater in telbivudine compared with lamivudine recipients at all study visits from week 12 to week 104 and the difference increased over time (data not shown). Among the 407 HBeAgpositive patients with HBV genotype C, HBeAg seroconversion occurred in significantly more telbivudine- than lamivudine-treated patients (32% vs 23%, respectively; P.0295). Among the 588 HBeAg-positive patients, 64% had a baseline ALT level at least twice the ULN, a population recommended for treatment by current guidelines. 8,9 HBeAg loss and HBeAg seroconversion occurred in significantly more telbivudine versus lamivudine recipients (42% vs 32% [P.021] and 36% vs 27% [P.022], respectively). Patients with baseline ALT levels at least twice the ULN also demonstrated higher rates of response on other end points compared with the overall HBeAg-positive patient population (Table 3). Among patients with baseline ALT levels 5 times the ULN (75 U/L for telbivudine and 86 U/L for lamivudine), HBeAg seroconversion was achieved by 46% of telbivudine recipients and 31% of lamivudine recipients (P.049). Very few patients in either treatment group experienced HBsAg loss or HBsAg seroconversion during the study. Among HBeAg-positive patients, 6 patients (1.3%) each receiving lamivudine and telbivudine had HBsAg loss by week 104; among these 6 patients, 3 (0.7%) receiving lamivudine and 2 (0.4%) receiving telbivudine had HBsAg seroconversion (HBsAg loss with detectable antibody to hepatitis B surface antigen) by week 104 (Table 2). Among HBeAg-negative patients, HBsAg loss was un-

5 490 LIAW ET AL GASTROENTEROLOGY Vol. 136, No. 2 Table 2. Efficacy of Telbivudine Versus Lamivudine at Week 104 (Intent-to-Treat Population) HBeAg-positive patients HBeAg-negative patients Parameter Telbivudine (n 458) Lamivudine (n 463) Difference (95% CI) P value Telbivudine (n 222) Lamivudine (n 224) Difference (95% CI) P value Therapeutic response (%) ( ) ( ).007 Nondetectable by PCR (%) ( ) ( ).001 HBV DNA 5 log ALT normalization (%) ( ) ( 0.7 to 16.2).073 HBeAg loss (%) ( 0.1 to 12.2).056 N/A N/A N/A N/A HBeAg seroconversion (%) ( 0.9 to 10.8).095 N/A N/A N/A N/A HBsAg loss (%) ( 1.5 to 1.5) ( 2.0 to 1.1).567 HBsAg seroconversion (%) ( 1.2 to 0.7) ( 1.2 to 1.2).995 Viral breakthrough a (%) Viral resistance b (%) NOTE. Missing values were considered as failure for categorical end points. CI, confidence interval; N/A, not applicable. a 1 log above nadir virologic breakthrough was defined as those patients with a confirmed response (2 consecutive 1 log 10 copies/ml HBV DNA decreases below baseline) who subsequently had a confirmed 1 log 10 copies/ml HBV DNA increase above nadir (the lowest HBV DNA level achieved). b Resistance is defined as viral breakthrough with treatment-emergent resistance mutations confirmed by genetic sequencing at week 104. In the study protocol, viral breakthrough was defined primarily as an increase in serum HBV DNA level to at least 5 log 10 copies/ml, following reduction to below that level. During the course of the study, this definition was superseded by the simpler and more widely accepted definition that pertains to the data reported above. Resistance rates per protocol were 17.8% and 30.1% for telbivudine and lamivudine, respectively, in HBeAg-positive patients, and 7.3% and 16.6% (P.001), respectively, in HBeAg-negative patients. Table 3. Week 104 Outcomes in HBeAg-Positive Patients With Baseline ALT Level 2 Times the ULN Telbivudine Lamivudine P value n Therapeutic response (%) HBV DNA decrease from baseline (mean log 10 ) HBV DNA nondetectable by PCR (%) HBV DNA 5 log ALT normalization ( times the ULN) (%) HBeAg loss (%) HBeAg seroconversion (%) Treatment failure a (%) a Treatment failure includes primary treatment failure (HBV DNA never 5 log) and resistance-related secondary treatment failure. changed from week 52 in 2 patients (0.9%) receiving lamivudine and 1 (0.5%) receiving telbivudine. One patient in each treatment group (0.4% for lamivudine and 0.5% for telbivudine) had HBsAg seroconversion by week 104. Posttreatment durability of HBeAg responses. By week 104, 119 HBeAg-positive telbivudine recipients (26%) and 102 lamivudine recipients (22%) met the criteria for treatment discontinuation due to efficacy (Table 4). A total of 39 telbivudine recipients and 20 lamivudine recipients discontinued treatment for efficacy at the discretion of the clinical investigator and continued to follow the study visit schedule. One lamivudine recipient and one telbivudine recipient who discontinued for efficacy were excluded from the analysis due to a lack of posttreatment follow-up data. Of the HBeAg-positive patients who discontinued treatment for efficacy, 31 of 38 telbivudine recipients (82%) and 17 of 19 lamivudine recipients (89%) sustained HBeAg loss through their last study visit. The majority of these patients also seroconverted to anti-hbe: 35 of 38 with HBeAg loss while receiving telbivudine and 17 of 19 while receiving lamivudine. The median time from HBe seroconversion to drug discontinuation was 40 weeks for telbivudine and 47 weeks for lamivudine. After treatment discontinuation, 28 of 35 telbivudine recipients (80%) Table 4. Week 104 Characteristics of HBeAg-Positive Patients Who Qualified for Treatment Discontinuation During the Study Telbivudine n/n (%) Lamivudine n/n (%) Patients discontinued treatment 39/119 (33) 20/102 (20) HBeAg loss 31/38 (82) 17/19 (89) HBeAg seroconversion 28/35 (80) 15/17 (88) HBV DNA suppression ( 5 log 10 ) 25/38 (66) 18/19 (95) HBV DNA nondetectable ( /38 (29) 8/19 (42) copies/ml) ALT normalization ( 1 times the 21/33 (64) 11/16 (69) ULN) Patients did not discontinue treatment 80/119 (67) 82/102 (80) HBeAg loss 69/80 (86) 68/82 (83) HBeAg seroconversion 60/80 (75) 61/82 (74) HBV DNA suppression ( 5 log 10 ) 71/80 (89) 73/82 (89) HBV DNA nondetectable ( /80 (85) 61/82 (74) copies/ml) ALT normalization ( 1 times the ULN) 69/80 (86) 73/82 (89)

6 February 2009 TELBIVUDINE VS LAMIVUDINE FOR HEPATITIS B 491 and 15 of 17 lamivudine recipients (88%) sustained HBeAg seroconversion. The median duration of posttreatment follow-up was 29.1 weeks and 32.6 weeks for telbivudine and lamivudine recipients, respectively. Viral breakthrough and resistance. Viral breakthrough and genotypic resistance were significantly less common with telbivudine than lamivudine (Table 2). At week 104, the signature mutation associated with telbivudine resistance in patients with viral breakthrough was M204I, found either alone or in association with the secondary mutations L80I/V or L180M. A single case of telbivudine resistance was associated with the M204V/ L180M double mutation at week 104. This was the first and only instance of this mutation during the study. Variants harboring mutations at the L229 codon were detected sporadically but were not associated with viral breakthrough without concomitant presence of M204I. Twenty-four telbivudine-treated and 42 lamivudinetreated patients who developed virologic breakthrough after a median of 60 (range, ) weeks were treated with adefovir dipivoxil 10 mg once daily for at least 16 weeks. At the discretion of the investigator, 19 of the telbivudine-treated and 31 of the lamivudine-treated patients continued a combination with add-on adefovir dipivoxil, while 5 of the telbivudine-treated and 11 of the lamivudine-treated patients were switched to adefovir dipivoxil monotherapy. For the patients treated with telbivudine, the mean change in serum HBV DNA level from the time of adefovir dipivoxil commencement through week 16 of therapy was 4.3 log 10 and 3.7 log 10 copies/ml in patients on combination therapy (n 19) and monotherapy (n 5), respectively. Two HBeAgpositive patients lost HBeAg while receiving adefovir dipivoxil follow-on treatment (one patient on combination therapy and one on monotherapy). The A181T mutation was not found in patients with viral breakthrough but was only observed in some patients with serum HBV DNA samples 3 log 10 copies/ml. After 48 weeks of telbivudine therapy, 15 patients with A181T mutations were identified; none of these developed virologic breakthrough during the second year while continuing telbivudine treatment. This mutation was less common at year 2 than year 1, because 8 patients reverted to wild-type A181 in year 2. A further 4 patients developed A181T changes (2 A181T; 2 A181A/T) during year 2, and none of these experienced virologic breakthrough. Analyses of baseline and on-treatment HBV DNA levels as predictors of 2-year efficacy end points. Multivariate logistic regression analyses identified treatment with telbivudine as a significant, independent predictor of better efficacy outcomes at week 104 for both HBeAgpositive patients (higher rates of PCR negativity, HBeAg seroconversion rate, and ALT normalization and less resistance) and for HBeAg-negative patients (higher PCR negativity rate and less resistance). A stepwise multivariate logistic regression was used to select significant baseline predictors for the week 104 efficacy outcome (P.25 to enter, P.1 to stay). Therefore, we analyzed in greater detail the impact of serum HBV DNA level at baseline and the early ontreatment virologic response (HBV DNA level at treatment week 24) to predict the efficacy outcomes after 2-year treatment with telbivudine. Impact of baseline HBV DNA level. In HBeAgpositive patients, serum HBV DNA levels 9 log 10 copies/ml or ALT levels 2.0 times the ULN at baseline were strong predictors for the virologic outcomes at week 104, including nondetectable serum HBV DNA by PCR (P.001 for both) and HBeAg seroconversion (P.002 and P.001, respectively; Figure 2). Baseline HBV DNA level was also a significant predictor of low resistance at week 104 (P.001) (odds ratio for baseline HBV DNA 9 vs 9 log 10 copies/ml, 0.28 [95% confidence interval, ]). Figure 2. Multivariate logistic regression analyses of the relationship of baseline virologic characteristics and type of treatment for week 104 outcomes for the intent-to-treat population. The figure shows the odds ratio (OR) 95% confidence interval (CI) of the parameters that are significant (P.001) predictors of week 104 outcomes. (A) Baseline variables in HBeAg-positive patients identified by multivariate analysis as significant predictors of outcomes after 104 weeks of telbivudine treatment. ORs 1 indicate direct relationships; ORs ratios 1 indicate inverse relationships. Selection criteria were as follows: baseline HBV DNA level 9vs 9 log 10 copies/ml; baseline ALT level 2.0 times the ULN vs 2.0 times the ULN. (B) Baseline variables in HBeAg-negative patients identified by multivariate analysis as significant predictors of outcomes after 104 weeks of telbivudine treatment. Data indicate ORs 95% CIs. ORs 1 indicate direct relationships; ORs 1 indicate inverse relationships. Selection criteria were as follows: baseline HBV DNA level 7vs 7 log 10 copies/ml; baseline ALT level 2.0 times the ULN vs 2.0 times the ULN; baseline BMI 24.5 vs 24.5 kg/m 2. BMI, body mass index.

7 492 LIAW ET AL GASTROENTEROLOGY Vol. 136, No. 2 Figure 3. Relationship between early virologic response to telbivudine and week 104 outcomes. Data indicate rates of efficacy outcomes and resistance at week 104 according to serum HBV DNA level achieved after 24 weeks of telbivudine treatment. Resistance rates at week 104, calculated using the protocol definition of viral breakthrough (persistent increase of HBV DNA to 5 log 10 copies/ml after suppression to below that level), were 4% and 3% for telbivudine-treated HBeAg-positive patients with ALT level 2 times the ULN and HBeAg-negative patients, respectively, who were PCR negative at week 24. Table 5. Frequency of Most Common Clinical Adverse Events at Week 104 (Occurring in at Least 5% of Patients in Either Treatment Group; Safety Population) Telbivudine (n 680) Lamivudine (n 687) Upper respiratory tract infection Nasopharyngitis Fatigue Headache Postprocedural pain Cough Influenza Diarrhea Abdominal pain upper Nausea Pharyngolaryngeal pain Arthralgia Dizziness Dyspepsia NOTE. All values are expressed as percentages. In HBeAg-negative patients, baseline HBV DNA level was not a statistically significant predictor of outcomes. The subgroup of patients (n 91) with baseline HBV DNA level 7 log 10 copies/ml showed better outcomes at week 104 in comparison with all HBeAg-negative patients, a higher rate of nondetectable viremia (89.2% vs 82.0%), and lower resistance (3.1% vs 10.8%, respectively). Early on-treatment virologic response correlates with week 104 outcomes. In addition to the analyses of baseline HBV DNA and ALT levels described previously, we performed a categorical analysis of serum HBV DNA levels after 12 and 24 weeks of telbivudine treatment. The proportions of patients with nondetectable HBV DNA increased markedly between treatment weeks 12 and 24 from 19% to 45% in HBeAg-positive patients and from 48% to 80% in HBeAg-negative patients. The outcomes at week 104 for HBeAg-positive patients were significantly different for all outcome measures, depending on the HBV DNA levels at treatment week 24, including the proportion of PCR-negative patients, HBeAg seroconversion, normalization of ALT level, and resistance rates. These differences at week 104 were particularly marked for patients who were PCR negative at week 24 compared with those in whom serum HBV DNA level remained 4 log 10 copies/ml at this time point (Figure 3). The early on-treatment response, with low HBV DNA levels at week 24, was associated with favorable 2-year efficacy outcomes, regardless of treatment assignment. However, a higher proportion of telbivudine-treated patients achieved nondetectable viremia at treatment week 24. In addition, telbivudine recipients with undetectable viremia at treatment week 24 achieved the highest rates of therapeutic response, PCR negativity, and HBeAg seroconversion and the lowest rates of resistance at week 104 (Figure 3). Conversely, patients with the least viral suppression at week 24 (HBV DNA level 4 log 10 copies/ml) had the lowest rates of efficacy and highest rates of resistance. Safety and Tolerability The proportion of patients reporting at least one adverse event through week 104, regardless of attributability to study drug, was similar for telbivudine and lamivudine (81% vs 77%, respectively). Clinical adverse events reported in more than 5% of the study population are detailed in Table 5. Most were constitutional symptoms, mild, transient, and not attributed to study drug. Adverse events considered by the investigator to be possibly related to study treatment were reported in 197 telbivudine recipients (29%) and 159 lamivudine recipients (23%). The proportion of patients with adverse events leading to treatment interruption or discontinuation was the same in both treatment groups (4%). Serious adverse events, regardless of attributability to study drug, were reported for 33 telbivudine recipients (5%) and 44 lamivudine recipients (6%). A total of 5 drug-related serious adverse events were reported in the study: 3 telbivudine recipients (myopathy, liver failure, and elevated CK level) and 2 lamivudine recipients (urticarial rash, hepatitis flare). Myopathy, characterized by muscle pain and weakness and moderately elevated CK levels during treatment, was reported in 2 patients; these cases were considered by the investigator to be related to telbivudine therapy. Both cases resolved after drug discontinuation. One patient receiving telbivudine had treatment interrupted for 45 days due to a drug-related serious adverse event (myopathy). As this patient achieved HBeAg seroconversion, telbivudine treatment was discontinued at week 52 due to efficacy. One patient developed liver failure at week 76

8 February 2009 TELBIVUDINE VS LAMIVUDINE FOR HEPATITIS B 493 (peak ALT level, 2350 IU/L; bilirubin level, 10.3 mg/dl) due to viral breakthrough during telbivudine treatment. The patient was discontinued from the protocol and managed at the discretion of the investigator with subsequent recovery. A total of 116 patients developed grade 3/4 CK elevations (7 times the ULN) through the duration of the GLOBE study; however, these did not correlate with musculoskeletal side effects and no cases of rhabdomyolysis were reported during this study. The grade 3/4 CK elevations were observed more frequently in patients receiving telbivudine (n 88; 12.9%) compared with lamivudine recipients (n 28; 4.1%) (P.001). The mean time to the first grade 3/4 CK elevation was 56.9 weeks for telbivudine-treated patients and 42.1 weeks for lamivudine-treated patients. The mean CK values ranged between 7.0 and 56.0 times the ULN for telbivudinetreated patients and between 7.1 and times the ULN for lamivudine-treated patients. There was no increase in MB fraction of CK. The majority of on-treatment grade 3/4 CK elevations were transient and resolved (decreased to grade 0 to 2) by the time of the next CK determination, a retest at an unscheduled or scheduled visit. Most patients with grade 3/4 CK elevations had only a single episode of a grade 3/4 CK elevation. At the time of grade 3/4 CK elevations, the mean serum ALT and aspartate aminotransferase (AST) levels for telbivudine-treated patients were 1.18 times and 2.13 times the ULN, whereas for lamivudine-treated patients these levels were 1.69 times and 2.31 times the ULN, respectively. During the GLOBE study, grade 3/4 ALT or AST elevations during treatment were less frequent with telbivudine compared with lamivudine (ALT 6.3% vs 11.6% and AST 6% vs 8.9%, respectively). Between treatment week 24 and the end of the study, ALT flares occurred less frequently in telbivudine recipients compared with lamivudine recipients (1.5% vs 4.8%). Discussion Two years of therapy with telbivudine was superior to lamivudine in both HBeAg-positive and HBeAgnegative patients for achieving therapeutic response, the primary efficacy end point. The intent-to-treat analysis at 2 years showed that telbivudine treatment was also superior to lamivudine for all direct measures of antiviral efficacy, including reduction in serum HBV DNA level from baseline, rates of PCR negativity, and viral resistance, consistent with the 1-year results from this study. 16 The differences between telbivudine and lamivudine treatment groups widened from year 1 to year 2 for all study end points in both HBeAg-positive and HBeAgnegative patients. Moreover, at week 104, telbivudine was statistically superior for therapeutic response in HBeAgnegative patients and for normalization of ALT level in HBeAg-positive patients, while at week 52 the results for both of these comparisons met only noninferiority criteria for telbivudine versus lamivudine. 16 The rates of HBeAg loss and HBeAg seroconversion also increased with prolonged treatment. In an earlier 2-year lamivudine trial involving 194 patients, the HBeAg seroconversion rates were 25% 27%. 27 Consistent with these findings, the present intent-to-treat analysis including 355 patients receiving lamivudine for 2 years showed very similar rates, with HBeAg loss and HBeAg seroconversion of 29.2% and 24.7%, respectively. However, the corresponding rates in 355 patients receiving 2-year therapy with telbivudine were higher (35.2% and 29.6%; P.056 and P.095, respectively). When the analysis was focused on the population recommended for treatment according to current clinical practice guidelines (baseline serum ALT level 2 times the ULN), the difference in the seroconversion rates between treatment with telbivudine (36%) and treatment with lamivudine (27%) was statistically significant (Table 3). The rates of HBeAg loss were significantly higher after 1 year of treatment with telbivudine than with lamivudine in the phase 3 registration trial in China, 28 thus supporting the conclusion that telbivudine therapy results in higher HBeAg loss and HBeAg seroconversion than lamivudine. A recent study comparing lamivudine with entecavir showed that the proportions of entecavir-treated and lamivudine-treated patients achieving HBeAg seroconversion were not significantly different through 96 weeks of treatment. 29 The durability of HBeAg seroconversion after discontinuation of telbivudine therapy was more than 80% and similar to that of lamivudine. According to the study protocol, discontinuation was possible after at least 24 additional weeks of treatment following initial detection of HBeAg loss. As previously shown, a period of continuous treatment to consolidate HBeAg seroconversion is important to sustain such a high rate after treatment discontinuation. 30 The durability of HBeAg seroconversion, as shown here in patients who discontinued therapy for efficacy, suggests that telbivudine therapy may be stopped in patients after HBeAg seroconversion, which is an important consideration in clinical management and cost-effectiveness assessments. Patients who discontinued treatment for efficacy showed lower rates of HBV suppression and ALT normalization than those who continued therapy with telbivudine or lamivudine after HBeAg seroconversion. This can be expected because the HBV control depends on host immunity, rather than pharmacologic suppression, and requires continued and regular patient monitoring. Similarly, temporary fluctuations in serum HBV DNA levels were observed after discontinuation of long-term adefovir treatment. 31 During the second year of treatment, the proportions of HBeAg-positive and -negative patients with undetectable serum HBV DNA decreased with both lamivudine and telbivudine, and fewer lamivudine recipients maintained PCR negativity throughout the study. The observed decline in

9 494 LIAW ET AL GASTROENTEROLOGY Vol. 136, No. 2 PCR negativity is largely due to development of drug resistance. All nucleos(t)ide therapies for chronic hepatitis B have shown various rates of loss of response due to drug resistance. 28,32 34 A direct comparison of reported resistance rates associated with different nucleos(t)ides is not possible because of differences in patient populations and study design. The head-to-head comparison in the present study clearly showed that the rate of telbivudine resistance is significantly lower than the lamivudine resistance rate. Importantly, the previously reported relationships between the magnitude of early (week 24) viral suppression and subsequent outcomes were found to persist through 2 years of treatment. Telbivudine recipients who were PCR negative at week 24 showed the highest rates of efficacy responses after 2 years of treatment, with resistance rates substantially lower than those in the overall population. Patients with higher levels of residual HBV DNA at week 24 had lower rates of efficacy responses at week 104 and higher rates of resistance. For these patients, modifying the treatment regimen will aim to enhance HBV suppression to achieve nondetectable serum HBV DNA and in the long-term to minimize viral breakthrough. Further studies are needed to determine the optimal strategy for adjusting therapy in this situation. Based on the cross-resistance profile, 35,36 add-on treatment with nucleotide analogues, such as adefovir or tenofovir, deserves prospective evaluation in future studies. Consistent with previous findings, the M204I signature mutation was the primary basis for telbivudine resistance, with secondary mutations detected at the L80, L180, and other codons. 16,21,35 In vitro studies have shown that HBV with the M204I mutation remains sensitive to the nucleotide analogues, adefovir dipivoxil and tenofovir. 35,36 This is supported by the results that we observed in 24 patients who developed telbivudine resistance in this trial; switch or add-on adefovir treatment reduced serum HBV DNA levels between 3.7 and 4.3 logs in 16 weeks, indicating that viral suppression can be restored by switch-to or add-on adefovir. The preliminary data in this small group of patients seem to favor the latter, similar to the accumulating evidence in patients with lamivudine resistance, which warrants further investigation in future studies. The A181T mutation was detected in several telbivudine recipients but was not associated with viral breakthrough and with continued telbivudine treatment was replaced by wild type in most patients. These results indicate that the A181T mutation is unlikely to be associated with reduced sensitivity to telbivudine. Both study treatments were well tolerated through week 104, consistent with previous clinical findings 16,21 and preclinical toxicology, 40 suggesting a satisfactory safety profile for extended telbivudine therapy. CK elevations were more frequent with telbivudine versus lamivudine, but these were not associated with musculoskeletal adverse events. The majority of grade 3/4 CK elevations were transient and resolved by the time of the next testing while the patients remained on treatment. There was no parallel increase in the MB fraction of CK or serum aminotransferase levels. The introduction of lamivudine in 1998 revolutionized our approach to the treatment of patients with chronic hepatitis B, and lamivudine remains a widely prescribed therapy worldwide. In the GLOBE study, telbivudine demonstrated superiority in comparison with lamivudine on all direct measures of antiviral efficacy, with an excellent safety and tolerability profile. Multivariate regression analyses identified telbivudine treatment as a significant independent predictor for better efficacy outcomes at week 104. Rates of HBeAg seroconversion with telbivudine after 1 and 2 years are higher than those reported with other agents, and the early antiviral effect of telbivudine appears greater than that obtained with adefovir and lamivudine and similar to that of entecavir. 13,14,16,22,29 Nondetectable serum HBV DNA at treatment week 24 was associated with most favorable 2-year efficacy outcomes. In summary, the final 2-year GLOBE trial results show that telbivudine has significantly greater 2-year efficacy compared with lamivudine, both in HBeAg-positive and HBeAg-negative patients, and is superior to lamivudine in achieving the treatment goals in patients with chronic hepatitis B. References 1. McMahon BJ. Epidemiology and natural history of hepatitis B. Semin Liver Dis 2005;25(Suppl 1): Lok AS, McMahon BJ. Chronic hepatitis B AASLD Practice Guidelines. Hepatology 2001;34: Fattovich G. Natural history of hepatitis B. J Hepatol 2003; 39(Suppl 1):S50 S Chen CJ, Yang HI, Su J, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA 2006;295: Iloeje UH, Yang HI, Jen CL, et al. Risk and predictors of mortality associated with chronic hepatitis B infection. Clin Gastroenterol Hepatol 2007;5: Iloeje UH, Yang HI, Su J, et al. Predicting cirrhosis risk based on the level of circulating hepatitis B viral load. Gastroenterology 2006;130: Chen G, Lin W, Shen F, et al. Past HBV viral load as predictor of mortality and morbidity from HCC and chronic liver disease in a prospective study. Am J Gastroenterol 2006;101: Liaw YF, Leung N, Guan R, et al. Asian-Pacific consensus statement on the management of chronic hepatitis B: a 2008 update. Hepatology Int 2008;2: Lok AS, McMahon BJ. Chronic hepatitis B. Hepatology 2007;45: Lai CL, Chien RN, Leung NW, et al. A one-year trial of lamivudine for chronic hepatitis B. Asia Hepatitis Lamivudine Study Group. N Engl J Med 1998;339: Tassopoulos NC, Volpes R, Pastore G, et al. Efficacy of lamivudine in patients with hepatitis B e antigen-negative/hepatitis B virus DNA-positive (precore mutant) chronic hepatitis B. Hepatology 1999;29: Hadziyannis SJ, Tassopoulos NC, Heathcote EJ, et al. Adefovir dipivoxil for the treatment of hepatitis B e antigen-negative chronic hepatitis B. N Engl J Med 2003;348: Marcellin P, Chang TT, Lim SG, et al. Adefovir dipivoxil for the treatment of hepatitis B e antigen-positive chronic hepatitis B. N Engl J Med 2003;348:

10 February 2009 TELBIVUDINE VS LAMIVUDINE FOR HEPATITIS B Chang TT, Gish RG, De MR, et al. A comparison of entecavir and lamivudine for HBeAg-positive chronic hepatitis B. N Engl J Med 2006;354: Lai CL, Shouval D, Lok AS, et al. Entecavir versus lamivudine for patients with HBeAg-negative chronic hepatitis B. N Engl J Med 2006;354: Lai C-L, Gane E, Liaw Y-F, et al. A comparison of telbivudine and lamivudine in patients with chronic hepatitis B. N Engl J Med 2007;357: Tenofovir EMEA approval. Available at: eu/humandocs/pdfs/epar/viread/h-419-pi-en.pdf. 18. Tenofovir FDA approval. Available at: fda.gov/scripts/cder/drugsatfda/index.cfm?fuseaction Search. DrugDetails. 19. Yuen MF, Sablon E, Hui CK, et al. Factors associated with hepatitis B virus DNA breakthrough in patients receiving prolonged lamivudine therapy. Hepatology 2001;34: Gauthier J, Bourne EJ, Lutz MW, et al. Quantitation of hepatitis B viremia and emergence of YMDD variants in patients with chronic hepatitis B treated with lamivudine. J Infect Dis 1999;180: Lai CL, Leung N, Teo EK, et al. A 1-year trial of telbivudine, lamivudine, and the combination in patients with hepatitis B e antigen-positive chronic hepatitis B. Gastroenterology 2005;129: Chan HL, Heathcote EJ, Marcellin P, et al. Treatment of hepatitis B e antigen-positive chronic hepatitis with telbivudine or adefovir: a randomized trial. Ann Intern Med 2007;147: Zollner B, Schafer P, Feucht HH, et al. Correlation of hepatitis B virus load with loss of e antigen and emerging drug-resistant variants during lamivudine therapy. J Med Virol 2001;65: Division of AIDS table for grading severity of adult adverse experiences. Bethesda, MD: National Institute of Allergy and Infectious Diseases, Locarnini S, Hatzakis A, Heathcote J, et al. Management of antiviral resistance in patients with chronic hepatitis B. Antivir Ther 2004;9: Agresti A, Hartzel J. Strategies for comparing treatments on a binary response with multi-centre data. Stat Med 2000;19: Liaw YF, Leung NWY, Chang TT, et al. Asian hepatitis lamivudine study group. Effects of extended lamivudine therapy in Asian patients with chronic hepatitis B. Gastroenterology 2000;119: Hou J, Yin Y-K, Xu D, et al. Telbivudine versus lamivudine in Chinese patients with chronic hepatitis B: results at 1 year of a randomized, double-blind trial. Hepatology 2008;47: Gish RG, Lok AS, Chang TT, et al. Entecavir therapy for up to 96 weeks in patients with HBeAg-positive chronic hepatitis B. Gastroenterology 2007;133: Chien RN, Yeh CT, Tsai SL, et al. Determinants for sustained HBeAg response to lamivudine therapy. Hepatology 2003;38: Handziyannis S, Sevastianos V, Rapti IN, et al. Sustained biochemical and virological remission after discontinuation of 4 to 5 years of adefovir dipivoxil (ADV) treatment in HBeAg negative chronic hepatitis B. Hepatology 2006;44(Suppl 1):231A. 32. Lai CL, Dienstag J, Schiff E, et al. Prevalence and clinical correlates of YMDD variants during lamivudine therapy for patients with chronic hepatitis B. Clin Infect Dis 2003;36: Hadziyannis SJ, Tassopoulos NC, Heathcote EJ, et al. Long-term therapy with adefovir dipivoxil for HBeAg-negative chronic hepatitis B for up to 5 years. Gastroenterology 2006;131: Colonno RJ, Rose R, Baldick CJ, et al. Entecavir resistance is rare in nucleoside naïve patients with hepatitis B. Hepatology 2006; 44: Standring DN, Seifer M, Patty A, et al. HBV resistance determination from the telbivudine GLOBE registration trial (abstr). J Hepatol 2006;44(Suppl 2):S Lok AS, Zoulim F, Locarnini S, et al. Antiviral drug-resistant HBV: standardization of nomenclature and assays and recommendations for management. Hepatology 2007;46: Rapti I, Dimou E, Mitsoula P, et al. Adding-on versus switching-to adefovir therapy in lamivudine-resistant HBeAg-negative chronic hepatitis B. Hepatology 2007;45: Ijaz S, Arnold C, Dervisevic S, et al. Dynamics of lamivudineresistant hepatitis B virus during adefovir monotherapy versus lamivudine plus adefovir combination therapy. J Med Virol 2008; 80: Liaw YF. Rescue therapy for lamivudine-resistant chronic hepatitis B: when and how? Hepatology 2007;45: Bridges EG. Telbivudine preclinical safety studies suggest minimal risk of chronic toxicity, reproductive toxicity or carcinogenicity (abstr). J Hepatol 2006;44:S147. Received August 29, Accepted October 9, Address requests for reprints to: Yun Fan Liaw, MD, Liver Research Unit, Chang Gung Memorial Hospital, Chang Gung University College of Medicine, Taipei, Taiwan. liveryfl@so-net.net.tw; fax: (886) The authors thank Richard Boehme, Bruce Belanger, Weibin Bao, Catherine Pajak, and Charles Koehne for their assistance during manuscript preparation. Supported by Idenix Pharmaceuticals, Inc and Novartis Pharma AG. The corresponding author had full access to all of the data and takes full responsibility for the veracity of the data and analysis. Statistical analyses were performed by the study sponsor and independently by Quartesian, LLC (Princeton, NJ). The authors disclose the following: Y. F.L. was a consultant for Bristol-Myers Squibb, GlaxoSmithKline, and Novartis and has received grant/research support from Bristol Myers Squibb, Idenix, Novartis, and Gilead. E.G. received consulting fees from Gilead, GlaxoSmith- Kline, Merck, and Novartis and honoraria from GlaxoSmithKline, Idenix, Novartis, and Roche. N.L. is a speaker for Bristol Myers Squibb, GlaxoSmithKline, Novartis, and Schering Plough. S.Z. received consulting fees and/or lecture from Bristol Myers Squibb, Gilead, Glaxo- SmithKline, Novartis, Roche, and Schering Plough. C.L.L. has received research grants from Idenix-Novartis. E.J.H. served on advisory board for Axcan Pharma, Gilead Sciences, Hoffman-LaRoche, Novartis, Schering-Plough; received research grants/funding from Axcan Pharma, Bristol-Myers-Squibb, Gilead Sciences, GlaxoSmithKline, Hoffman-LaRoche, Idenix, Novartis, Pharmasset, Schering-Plough, Vertex. M.M. received research grants, honoraria, and/or consultancy fees from Bristol Myers Squibb, Boehringer Ingelheim, Gilead, Idenix, Novartis, Roche, Schering Plough, GSK, and Valeant. N.B. has received funding from Roche, Vertex, Schering, Gilead, Pharmasset, Connatus, Idenix, Novartis and speaker bureau fees from Gilead and Novartis. S. H.H. received grant support, consultancy fees, and lecture fees from Novartis/Idenix. S.G.H. received consulting fees from Bukwang and research grants from Idenix, Valeant, BMS, and GSK. G.P. serves on advisory boards for Bristol Myers Squibb, Gilead Sciences, Hoffmann La Roche, Idenix Novartis, has received research grants (from participation in phase 2/3 trials) from Gilead Sciences and Idenix Novartis, has received unrestricted research grants from Hoffmann La Roche, and is a speaker for Bristol Myers Squibb, Gilead Sciences, Hoffmann La Roche, and Idenix Novartis. N.A.B. and K.G. were employees of Idenix at the time of this study. E.A. was an employee of Novartis at the time of the study. N.V.N. is an employee of Novartis. J.N., M.J.T., Y.W., Y. Cakaloglu, and Y. Chen report no conflicts of interest.