This data & information was presented in a poster at EASL, April 2005, Paris by: JL Dienstag1, LLJ Wei2, D Xu3, AP Cross3, B Kreter3, RB Wilbur3 1Mass General Hospital & Harvard Medical School, Boston, MA; 2 Harvard School of Public Health, Boston, MA; 3Bristol Myers Squibb, Pharmaceutical Research Institute, Wallingford, CT & Princeton, NJ

Preceding the report from EASL here are links to additional reports & studies on entecavir:

FDA Approves New Treatment Baraclude (entecavir) for Chronic Hepatitis B URL: http://www.natap.org/2005/HBV/033105_01.htm - 31 Mar 2005

FDA Antiviral Drug Advisory Committee Recommends Unanimously to Approve Entecavir for Hepatitis URL: http://www.natap.org/2005/HBV/031505_02.htm - 15 Mar 2005

"Entecavir in HIV/HBV Co-Infected Patients: Safety and Efficacy in a Phase II Study (ETV-038)" URL: http://www.natap.org/2005/CROI/croi_18.htm - 08 Mar 2005

Entecavir: new hepatitis B drug-phase III study results of 48-weeks therapy in HBeAg+ nucleoside URL: http://www.natap.org/2004/AASLD/aasld_02.htm - 04 Nov 2004

Entecavir: new HBV drug, phase III study in nucleoside-naïve HBeAg-negative patients URL: http://www.natap.org/2004/AASLD/aasld_05.htm - 04 Nov 2004

URL: http://www.natap.org/2004/AASLD/aasld_30.htm - 16 Nov 2004

--lamivudine (LAM) & adefovir )ADV) were approved for the treatment of chronic hepatitis B based on results from placebo-controlled trials

--entecavit (ETV) is a new oral antiviral agent that has been found to be superior to LAM in phase II/III clinical trials in adults with chronic hepatitis B (CHB)

--the efficacy of ADV has not been studied relative to ETV or LAM in adequate & well controlled trials in adults with CHB

--the relative treatment efficacies of ETV, LAM & ADV are unknown

Compare data from the phase III clinical development program to combined efficacy from publicly available clinical trials of lamivudine & adefovir in adults with chronic hepatitis B infection, following a prospectively defined analysis protocol that applied statistical techniques.

(see details of results & methods used below)

--ETV was superior to LAM for all endpoints except ranked assessment of necroinflammation & ranked assessment of fibrosis, where the two agents are comparable

--ETV was superior to ADV for the virological endpoints & comparable to ADV for all other endpoints

--LAM was superior to ADV for the virological endpoints & comparable to ADV for all other endpoints

--ETV, ADV & LAM were all superior to placebo for all endpoints

The analysis suggest that at least two factors are important in comparisons among available & emerging anti-HBV therapies:

--the relative clinical activity of the antiviral agent, was consistent with results of observed potency from in vitro test systems

--characteristics of the virus detected in the patient (HBeAg+, HBeAg-)

--ETV was superior to LAM for virological endpoints & for ALT normalization both HBeAg subpopulations

--ETV was superior to ADV for virological endpoints in both HBeAg populations & for ALT normalization & seroconversion in the HBeAg-positive subpopulations

--ETV was superior to LAM for Histological Improvement & comparable to LAM for ranked assessment of necroinflammation & ranked assessment of fibrosis in both subpopulations

--ETV was superior to ADV for histological endpoints in the HBeAg-positive subpopulations & comparable in the HbeAg-negative subpopulations

--the consistent rankings of ETV>=LAM>=ADV across multiple endpoints supports the choice of ETV for initial therapy of HBV in nucleoside-naive subjects

--all antiviral therapies were consistently superior to placebo for all endpoints

Note from Jules Levin: long-term comparative study & therapy sequencing are needed to provide guidance for clinicians. As well, as we begin to study combination therapy & new drugs for HBV, more information will become available for guidance.

An analysis plan was prospectively developed to describe the process for identifying, reviewing, evaluating, and integrating the clinical trials literature relating to the use of ETV, LAM, & ADV for the treatment of CHB in adults.

A comprehensive search of electronic databases (MEDLINE, EMBASE, Intl Pharmaceuticals Abstracts), Biosis Previews, Derwent Drug File) was conducted on 27 May 2004, using Dialog with no date restrictions.

All available abstracts, published 1999-2004 inclusive, for 3 international liver meetings (EASL, AASLD, & APASL), were individually searched for relevant abstracts.

Primary data from FDA Summary Basis of Approval documents were identified by searching www.fda.gov

Relevant articles were identified & filtered, in a 2 stage process, using prospectively-defined inclusion & bexclusion criteria.

The inclusion criteria included 2 study types:

--prospective, controlled, randomized, human clinical trials that included at least one antiviral monotherapy arm & a control arm.

--prospectively-followed case series that included at least 15 patients receiving antiviral monotherapy

Eligible studies were those that included patients with all the following characteristics at baseline:

--HBsAg-positive for at least 24 weeks

--established CHB stratified by HBeAg status, prior nucleoside/nucleotide exposure & HBV DNA viremia

--serum ALT levels at baseline of 1.0 to 10 times the upper limit of noral

--compensated liver disease

--monotherapy with ETV (o.5 mg/day), ADV (10mg/day) or LAM mg/day)

Eligible studies were those that reported data for at least one of the endpoints assessed in this cross-study analysis.

The key exclusion criteria, and the numbers of articles selected & eliminated at each stage are summarized;

19 papers not in English

20 non-clinical studies

31 studies including patients <16 yrs old

78 studies were not clinical trials or did not meet criteria in protocol

3 studies did not include patients with CHB

7 no data for ETV, ADV, or LAM monotherapy

1 patient population studied was not identified as nucleoside-naïve, and without significant co-morbidity or co-infection (eg, HIV, HCV, HDV, post-liver transplant patients)

1 data determined not to be useful (other endpoint analyzed or treatment duration/dose varied from protocol)

14 duplicate publications

1 paper not in English

1 study includingf patients <16 yrs old

66 studies were not clinical trials or did not meet criteria in protocol

43 no data for ETV, ADV, or LAM monotherapy

76 patient population studied was not identified as nucleoside-naïve, and without significant co-morbidity or co-infection (eg, HIV, HCV, HDV, post-liver transplant patients)

8 the patient population was not described in a way as to allow for data extraction

74 data determined not to be useful (other endpoints analyzed or treatment duration/dose varied from protocol)

32 data was unavailable

65 duplicate publications

5 study included LAM-refractory patients

Two independent individuals extracted data from the relevant studies of interest onto a Microsoft Excel spreadsheet containing predefined fields;

--unique reference identifier, source author & reference, treatment groups, study design, number of patients, baseline demographics, baseline disease characteristics, histological results, virological results, serological results, ALT normalization, deaths

At the outset, patients were stratified into two categories:

--HBeAg positive, nucleoside-naïve patients with presumed wild-type virus

--HBeAg-negative, nucleoside-naïve patients with presumed wild-type virus

The endpoints were those collected at 1 year (defined as 48 to 52 weeks of treatment)

6 binary outcomes were assessed:

--histological improvement (proportion of patients with a >2 point decrease in the Knodell necro-inflammatoryscore from baseline and no worsening of fibrosis, defined as a >=1-point increase in the Knodell fibrosis score from baseline

--ranked assessment of necroinflammation (among patients with paired biopsy specimens, the proportion with less inflammation after 1 yr compared to baseline)

--ranked assessment of fibrosis (among patients with paired biopsy specimens, the proportion with less fibrosis after 1 year compared with baseline)

--HBV DNA <LOQ by PCR assay (proportion of patients with HBV DNA <=400 copies/ml and the LOQ used in the trial)

--HBV DNA <LOQ by non-PCR assays (proportion of patients with HBV DNA level less than the LOQ used in the trial)

--ALT normalization (proportion of patients with ALT <1 x ULN)

Two continuous outcomes were assessed:

--HBV DNA reduction from baseline by PCR assay (mean reduction from baseline in HBV DNA level)

--for HBeAg-positive patients only, seroconversion (proportion of patients with loss of HBeAg & acquisition of HBeAb)

These study references were used to compare the drugs:

ETV study 022

LAM study 022

ADV Marcellin 2003

PBO Marcellin 2003

LAM consensus (2 references)

PBO consensus (2 references)

--meta-analysis techniques were used to construct pooled (consensus) estimates of treatment effects for LAM, ADV, & placebo for continuos & binary endpoints

--treatment effect estimates were pooled from individual study results based on a fixed effects model

--for each endpoint, comparisons of individual treatment effects & consensus treatment effects across & within studies were based on the Student’s t-test to determine a relative ordering within the HBeAg-positive & HBeAg-negative subpopulation

--for binary endpoints, a test for homogeneity (ch-square) was performed for treatment arms contributing to a consensus estimate

ETV>ADV

ETV>LAM

LAM=ADV

ETV, LAM, ADV>PBO

(data not shown)

ETV>ADV

ETV=LAM

LAM=ADV

(data not shown)

ETV>ADV

ETV=LAM

LAM>ADV

ETV, LAM, ADV>PBO

ETV>ADV

ETV>LAM

LAM>ADV

ETV, LAM, ADV>PBO

ETV>ADV

ETV>LAM

LAM>ADV

ETV, LAM, ADV >PBO

(data not shown)

ETV>LAM

ETV, LAM>PBO

(data not shown)

ETV>ADV

ETV>LAM

LAM>ADV

ETV, LAM, ADV>PBO

ETV>ADV

ETV>LAM

LAM>ADV

ETV, LAM, ADV>PBO

ETV=ADV

ETV>LAM

LAM=ADV

ETV, LAM, ADV>PBO

(data not shown)

ETV=ADV

ETV=LAM

ETV, LAM, ADV>PBO

(data not shown)

ETV=ADV

ETV=LAM

LAM=ADV

ETV, LAM, ADV>PBO

ETV>ADV

ETV>LAM

LAM>ADV

ETV, LAM, ADV>PBO

ETV>ADV

ETV>LAM

LAM>ADV

ETV, LAM, ADV >PBO

(data not shown)

ETV>LAM

(data not shown)

ETV=ADV

ETV>LAM

LAM=ADV

ETV, LAM, ADV>PBO

ETV>ADV

ETV>LAM

LAM>ADV

ETV, LAM, ADV>PBO