HOME
Return to Table of Contents
276 NONINVASIVE ASSESSMENT OF FIBROSIS IN PATIENTS WITH CHRONIC HCV: A SYSTEMATIC REVIEW OF FIBROTEST® AND FIBROSCAN® AM Shaheen, AF Wan, RP Myers INTRODUCTION: Although biopsy remains the gold standard for assessing liver fibrosis, the development of noninvasive alternatives has emerged as a research priority. The most widely studied measures, the FibroTest® and FibroScan®, have gained widespread use predominantly due to their noninvasive nature and relative ease. The ideal measure has yet to be determined. Our objective was to determine the accuracy of these tests for the prediction of fibrosis in patients with chronic HCV.
Liver Unit, University of Calgary, Calgary, Alberta
METHODS: MEDLINE, EMBASE and Cochrane databases were searched (01/1997-10/2006) for studies assessing FibroTest® or FibroScan® for staging fibrosis in patients with chronic HCV. Studies with <30 subjects (n=1) or lacking sufficient data to assess test performance (n=4 for FibroTest®, n=5 for FibroScan®) were excluded. Random- and fixed-effects, summary receiver operating characteristic (SROC) curves were used to assess test performance according to the presence of heterogeneity between the studies.
RESULTS: 9 studies assessed the FibroTest® in 1,679 patients (60% male, mean age 46 yrs) and 4 studies assessed the FibroScan® in 546 patients (61% male, mean age 48 yrs). In the FibroTest® studies, the prevalence of significant fibrosis (METAVIR =>F2) and cirrhosis (F4) were 45% and 9%, respectively. For the FibroTest®, the areas under the SROC curves (AUCs) for significant fibrosis and cirrhosis were 0.79 (95% CI:0.77-0.81; 8 studies) and 0.90 (2 studies), respectively. In the FibroScan® studies, the prevalence of significant fibrosis and cirrhosis were 63% and 21%, respectively. The AUCs for significant fibrosis and cirrhosis were 0.82 (0.74-0.88; 4 studies) and 0.96 (0.85-0.99; 3 studies) for the FibroScan®.
CONCLUSION: Both the FibroTest® and FibroScan® have acceptable accuracy for the prediction of significant fibrosis and are highly accurate for cirrhosis in patients with chronic HCV. Based on this indirect comparison, the optimal measure is unclear, although other factors including cost and local availability need to be considered.