Objective: To describe the statistical association between serum AST and liver biopsy grade in patients with rheumatoid arthritis.
Methods: 94 patients from 3 prospectively followed cohorts underwent a total of 354 biopsies graded according to Roenigk. Blood samples for serum aminotransferase (AST) were obtained every 37.7 + or - 32.7 days (mean + or - SD) and 15.2 + or - 7.1 samples were obtained before each biopsy. For each prebiopsy interval, 4 AST functions were calculated: (1) mean, (2) maximum, (3) percentage abnormal, and (4) the presence or absence of at least one abnormality. Analysis of variance was performed to determine the effect of each of these variables on liver biopsy score.
Results: AST increased across biopsy grades: 26.5 IU + or - 10.7 IU (mean + or - SD) in Grade I biopsies, 28.4 + or - 10.9 in Grade II, and 35.4 + or - 21.1 in Grade IIIA (p = 0.0006, overall difference between classes). The percentage of abnormal prebiopsy AST values increased across biopsy grades: 8.7 + or - 13.9 (mean + or - SD) in Grade I biopsies, 12.3 + or - 17.5 in Grade II, and 18.6 + or - 27.1 in Grade IIIA samples [(p = 0.0014) overall difference between classes.] A mean prebiopsy AST in the abnormal range was more likely to be associated with a more abnormal liver biopsy grade (p = 0.01, Wilcoxon's rank sum test). AST values abnormal <49% of the time had a 97% specificity for a normal biopsy grade.
Conclusion: Regular AST measurements are useful markers of hepatic histologic outcome, within the range of mostly normal histology reported here, in patients with RA receiving longterm weekly MTX.