430A AASLD ABSTRACTS HEPATOLOGY, October 2003 559 AN OVERVIEW OF BIOCHEMICAL MARKERS (FIBROTEST- ACTITEST) DIAGNOSTIC VALUE IN CHRONIC LIVER DISEASES: A NON-INVASIVE ALTERNATIVE TO LIVER BIOPSY. Thierry Poynard, Fran~oise Imbert-Bismut, Vlad Ratziu, Groupe Hospitalier lh'tqe-Salpetfiere, Paris, France; Sylvie Naveau, H6pital Antoine Bdcl~re, Clamart, France; Dominique Thabut, Groupe Hospitalier lh'tqe-Salpetfiere, Paris, France; Didier Lebrec, H6pital Beaujon, Clichy, France; Philippe Halfon, RBML, Marseille, France; Fabien Zoulim, H6pital de l'H6tel Dieu, Lyon, France; Marc Bourliere, H6pital Saint Joseph, Marseille, France; Djamila Messous, Vincent Thibaut, Mona Munteanu, Groupe Hospitalier Pitqe-Salpetfiere, Paris, France Background: Liver biopsy is still the gold standard in chronic liver disease, but is invasive, costly ($1,000), has a high sampling error (33% for fibrosis and 24% for activity), as well as discordance rates greater than 10% between pathologists for routine scoring sys- tems. FibroTest-ActiTest (FF-AT) are biochemical markers of fi- brosis and activity ($90). The aim was to demonstrate, through an overview of diagnostic studies, that FF-AT were cost-effective enough to be used as non invasive alternatives to liver biopsy in patients with chronic hepatitis C (CHC), B (CHB), alcoholic liver disease (ALD) and non-alcoholic fatty liver disease (NAFLD). Methods: The following diagnostic parameters were assessed: the area under the ROC curves (AUROC), accuracy (AC), kappa sta- tistics (K), and Spearman correlation coefficient (SC). The main end point was the AUROC for the diagnosis of bridging fibrosis (F2F3F4 vs FOF1) and for the diagnosis of moderate/severe necro- sis activity (A2A3 vs AOA1). In order to assess the diagnostic value between each stage and grade, all possible combinations of the AUROC were computed. The quality of biopsy was scored accord- ing to published criteria: a single fragment, size of 15 mm or more, and 5 or more portal tracts. Specificity and normal values were determined prospectively in blood donors. The prevalence of the abnormal values (outside the 99% percentiles) of the 6 FT-AT components was assessed on 8,540 consecutive tests computed according to standardized procedures between September 2002 and May 2003. Results: A total of 10 studies were included with a total of 2,706 individual data: 2,407 subjects with both FF-AT and biopsy (1637 HCV, 127 HCV-HIV, 318 HBV, 228 ALD, 89 NAFLD, 8 others) and 299 healthy blood donors (BD). Mean age was 44 years (se-0.2), and 64% were males. Mean biopsy size was 16 mm (0.2)with 7 portal tracts (0.2) and 2.2 fragments (0.08). F2F3F4 was present at biopsy in 38% (cirrhosis 11%) and AZA3 in 37%. The mean FT values were linearly associated with fibrosis stages: 0.10 (se 0.01) in BD (n-299); 0.22 (0.01) in F0 (n-261); 0.33 (0.01) in F1 (n-1110); 0.48 (0.01) in F2 (n-497); 0.62 (0.03) in F3 (n-240); and 0.76 (0.01) in ]?4 (n-299) (p <0.001 between all groups by Dunnett's multicomparison test), (Figure 1). The diagnostic parameters for FF were: F2F3F4 vs FOF1; AUROC- 0.82 (0.01); AC- 75% (p<0.001); kappa- 0.47 (0.02); SC- 0.65 (p<0.001). For the diag- nosis of F4 vs FOFIF2F3, the AUROC was 0.88 (0.01). AT was associated with activity grades; The mean AT values were associ- ated with necrosis grades 0.09 (0.01) in BD (n-299); 0.28 (0.01) in A0 (n-516); 0.37 (0.01) in A1 (n-887); 0.55 (0.01) in A2 (n-600); and 0.59 (0.01) in A3 (n-404) (p<0.01 between all groups by Dunnett's multicomparison test). Diagnostic parameters for AT were AU- ROC-0.78 (0.01); AC-70% (p<0.001); K-0.41 (0.02); SC-0.54 (p<0.001). There were no gray zones between stages or grades, with similar AUROC between contiguous stages or grades: F0vsF1, FlvsF2, F2vsF3, F3vsF4. Factors associated with discordant results ->2 stages-grades between biopsy and FT-AT were the presence of less than 5 portal tracts (p-0.005) for FT and AT, and size being smaller than 15 mm for AT. The most frequent abnor- mal value observed during post-marketing follow-up was hapto- globin lower than 0.12 g/L (4.7% out of 8,540 tests), suggesting hemolysis as a possible cause of false positive FF-AT. Conclusion: Due to the cost and risk involved in liver biopsy, FibroTest-ActiTest are a cost-effective alternative in patients with chronic hepatitis C, B, ALD and NAFLD. Discordances observed were associated with small biopsy sample size and hemolysis for FF-AT. Q :.>. .......... !!!!i!:i ! i ! i ! i Z 2 ................................ " !.:.:.:.C Do~(~r FO F~ F2 Fa ~4 Disclosures: Marc Bourliere - No relationships to disclose Philippe Halfon - No relationships to disclose Fran~oise Imbert-Bismut - No relationships to disclose Didier Lebrec - No relationships to disclose Djamila Messous - No relationships to disclose Mona Munteanu - Biopredictive: Investigator Sylvie Naveau - No relationships to disclose Thierry Poynard - Biopredictive: Consultant/Advisor; Investigator Vlad Ratziu - No relationships to disclose Dominique Thabut - No relationships to disclose Vincent Thibaut - No relationships to disclose Fabien Zoulim - No relationships to disclose 560 MONITORING HCV CORE ANTIGEN DURING ANTIVIRAL THERAPY. Adrian M Di Bisceglie, Janice Stfinko, Patficia Osmack, Saint Louis University Health Sciences Center, SL Louis, MO Background: Measurement of HCV RNA is commonly used is assessing patients with chronic hepatitis C and their response to antiviral therapy. A novel assay is able to measure HCV core antigen in serum of infected patients. Aims: To assess the value of measuring HCV core antigen levels at baseline and during antiviral therapy. Methods: We measured HCV core antigen (Track-C, Ortho Clin- ical Diagnostics, Raritan NJ) in a group of HCV RNA positive patients enrolled in a randomized, controlled trial of pegylated interferon alfa-2a (180 mcg/wk) with or without ribavirin (1000- 1200 mg/day) in IFN naive patients infected with HCV genotype 1. HCV core antigen levels were measured before treatment, at days 1, 3, 7, 14, 22, 29 and at weeks 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48 and at follow up weeks 4, 12, and 24 after therapy using the Ortho assay. Results were compared to HCV levels at the same time points measured using the Roche Monitor assay. Results: 29 patients were enrolled of whom 28 have completed treatment and follow up. Mean HCV RNA level was 463,586 IU/ml (range 119,000 - 1,220,001). 24 were infected with HCV genotype la and 5 with genotype lb. Mean baseline HCV core antigen was 87.3 pg/ml (range 0.3-212.3). There was no correlation between HCV core Ag and ALT or AST levels. All but one patient was seropos- itive for HCV core antigen (>1.5 pg/ml) at baseline. There was a strong positive correlation between HCV RNA and HCV core Ag levels (r-0.907, p<0.001). So far 16 pts have had an end of treat- ment virologic response (ETR) and 9 a sustained virologic re- sponse (SVR) as assessed by HCV RNA. Levels of HCV core Ag decreased promptly on treatment in most patients and became negative in 21 of 29 pts at some time (mean time to first negative was 27.9 days, range I to 84 days). Of those who became negative, 16 had an ETR, 10 an SVR (1 pending) while 5 did not achieve a virologic response. There was no significant difference in mean HCV core antigen levels at baseline between those who achieved an SVR and those who did not (64.7 vs 101.6 pg/ml). All patients