Khalil G. Ghanem, Michael Torbenson, et al.

NIH Consensus Development Conference on
Management of Hepatitis C: 2002 

Bethesda, Maryland
June 10-12, 2002

Introduction

Liver biopsies are frequently recommended in the management of patients with chronic hepatitis C. Histologic criteria have been established to assess the severity of both inflammation and fibrosis. However, it remains uncertain how this information assists in establishing prognosis or predicting efficacy of treatment.

Objective

We conducted a systematic review of the literature to determine: (1) how the results of initial liver biopsy relate to measures of disease progression and treatment outcome as assessed by histologic and virologic parameters and (2) the value of serum biochemistry tests and serologic measures of fibrosis in predicting histologic findings.

Methods

Literature Sources

Seven electronic databases were searched through DIALOG for the period from January 1996 to March 2002. Additional articles were identified by searching references in pertinent articles, hand searching relevant journals, and querying technical experts.

Eligibility Criteria

Exclusion criteria for review included: non-English language, articles limited to basic science or non-human data, previously reported data, and meeting abstracts. Inclusion criteria for review were: study designed to address our key question, information pertinent to management of hepatitis C, and 30 or more study subjects with hepatitis

C. In addition, for those studies pertaining to how results of initial liver biopsy relate to measures of disease progression and treatment outcome, we required at least six months of followup after initial biopsy and outcomes measured by an appropriate objective standard such as virologic or histologic measures.

Assessment of Study Quality

Each eligible article was reviewed by a pair of reviewers, including at least one team  member with relevant clinical training and/or one with training in epidemiology and research methods. Paired reviewers  independently rated the quality of each study in terms of the following categories: representativeness of study subjects (5 items); bias and confounding (4 items); description of therapy (4 items); outcomes and followup (5 items); statistical quality and interpretation (4 items). Reviewers assigned each response level a score of 0 (criterion not met), 1 (criterion partially met), or 2 (criterion fully met) to each relevant item on the quality form.

The score for each category of study quality was the percentage of the total points available in each category and therefore could range from 0–100 percent. The overall quality score was the average of the five categorical scores. We also documented source of funding.

Extraction of Data

The paired reviewers also abstracted data on type of study and geographical location; the study groups; specific aims; the inclusion and exclusion criteria; demographic, social, and clinical characteristics of subjects; and results. Differences between the two reviewers in either quality or content abstraction were resolved by consensus.

Synthesis

Results of Literature Search

We identified 3,104 potentially relevant citations and 1,731 of these were eligible for abstract review. Through the abstract review process we identified 254 articles that could contain data on one of our key questions regarding the utility of liver biopsy in patients with chronic hepatitis C. After reviewing these 254 articles, we found 147 studies that addressed the value of initial or followup biopsies predicting treatment outcomes and 107 articles that addressed the relationship between serological markers and histological findings. We subsequently reviewed the full articles to ensure they met our eligibility criteria. We have focused on randomized controlled trials of therapies for which assignment to a treatment was not determined by biopsy results. Data from these eligible studies will be presented in a series of evidence tables and figures highlighting their distinguishing characteristics, methodologic strengths and limitations, and key findings.