Evidence for a Pathophysiological Role of Anti-ADAMTS13 Antibodies Despite the Presence of Normal ADAMTS13 Activity and Presumption of an Epitope Spreading over Time in Recurrent Thrombotic Thrombocytopenic Purpura (TTP).

Background: Only about 60% of patients with the clinical diagnosis of acute TTP display a severe ADAMTS13 deficiency. Consequently, it is of interest to know whether ADAMTS13 is implicated in the remaining 40%.

Materials and Methods: The patient is a currently 49-year-old male who suffered from 5 bouts of acute acquired TTP between 1998–2003. For each bout, he was treated with plasma exchange (PE) and achieved remissions after a median of 13 PE sessions (range 6–22). Diagnostic work-up during the first bout documented that he was HIV positive and although HIV medication was begun, he did not maintain follow-up appointments or continue his medication. Serum samples withdrawn before initiation of any form of plasma therapy were available for study from all episodes but the second. ADAMTS13 activity was determined by three different methods: quantitative immunoblotting (IB), a slightly modified FRETS-VWF73 assay, and a flow-based assay quantifying cleavage of unusually large VWF multimer strings. The presence of ADAMTS13 antibodies was investigated by a functional assay (IB), a commercial ELISA (TECHNOZYM ADAMTS13 INH ELISA), dot-immunobinding and by immune-precipitation experiments (epitope mapping) using recombinant ADAMTS13 fragments as well as an ADAMTS13 hybrid containing the ADAMTS1 spacer domain. In vivo proteolysis of endogenous VWF multimers was assessed using SDS-agarose electrophoresis.

Results: ADAMTS13 autoantibodies were documented in all episodes by ELISA and dot-immunobinding assay, which were associated with inhibitory capacity only in the fifth episode. Using the static IB assay, ADAMTS13 activity was normal in episodes 1 and 3, and only in his 5^th acute bout a severe ADAMTS13 deficiency (<5% of the normal) was documented. Using the flow-based assay, which closely mimicks the in vivo situation, a normal ADAMTS13 activity was found only during the first bout, which was at the same time the only episode with a normal in vivo VWF proteolysis. Fluorescence evolution in the modified FRETS-VWF73 assay followed a straight line over 3 hours in episodes 1 and 5, while progressingly increasing activities were recorded in episodes 3 and 4, explaining in part the observed discrepancies in ADAMTS13 activity. Epitope mapping revealed that in all episodes the principal anti-ADAMTS13 antibodies were directed against the ADAMTS13 spacer domain.

Conclusions: For the first time we have proven that ADAMTS13 is pathophysiologically involved in clinically diagnosed TTP despite normal ADAMTS13 activity in routine assays. Furthermore, an epitope spreading of the ADAMTS13 autoantibodies over time is likely. At each bout, their principal antigenic determinant was localized on the ADAMTS13 spacer domain and fine mapping is required to characterize their different functional effects.