Twenty-three lots of five antithrombin III (AT III) concentrates from
four manufacturers were analyzed in a single-blind study. Al the prepa
rations had been virus-inactivated by pasteurization, and one concentr
ate had also been treated with solvent/detergent (S/D). AT III activit
ies were determined using two thrombin-based and one factor Xa-based c
hromogenic substrate assays. AT III antigen was measured by kinetic ne
phelometry. All AT III assays were tested against the first internatio
nal reference preparation coded 72/1. In addition, AT III was characte
rized by crossed immunoelectrophoresis in the presence of heparin and
by gel filtration. The following were quantified: heparin cofactor II
activity and antigen content, heparin activity, thrombin-AT III comple
xes, AT III-protease complexes, total protein, albumin, immunoglobulin
s, glucose and pH. The AT III concentrates differed markedly in terms
of their purity and potency. The specific activities of AT III and the
ratios of AT III activity to antigen content ranged from 3.4 to 6.9 a
nd from 0.63 to 0.84, respectively. The highest values were found in f
ive lots of the concentrate that had been treated by both pasteurizati
on and S/D. This preparation was the only one that was virtually free
of denaturated AT III, as judged by crossed immunoelectrophoresis. Mar
ked batch-to-batch variation in AT III potencies was found in two out
of the five preparations analyzed. In two out of five lots from one ma
nufacturer, the measured potencies were more than 10% lower than the d
eclared potencies. None of the analyzed preparations contained heparin
cofactor II, and heparin activities were very low in all lots.