“…Stinton et al 16) determined the prevalences of different autoantibodies among 109 patients with PBC (including 6 patients negative for AMA) using LIA (82% for AMA -M2, 73% for M2 -3E, 27% for anti -Sp100 antibody, 17% for anti -PML antibody, 27% for anti -gp210 antibody, 32% for anti -Ro -52 antibody and 11% for ACA), which were different from those observed in the present study (65% for AMA -M2, 77.5% for M2 -3E, 13.8% for anti -Sp100 antibody, 8.7% for anti -PML antibody, 40.0% for anti -gp210 antibody, 27.5% for anti -Ro -52 antibody and 32.5% for ACA). The reported prevalences of these autoantibodies vary among studies : 75 -96% for AMA -M2 17) , 57% for M2 -3E 18) , 20 -30% for anti -Sp100 antibody 17,19,20) , 19% for anti -PML antibody 21) , 17 -25% for antigp210 antibody 17,22) , 25 -28% for anti -Ro -52 antibody 23,24) , and 10 -30% for ACA 25 -27) . This variability may reflect differences in target populations, such as racial difference, timing of sample collection and other variables among institutions.…”