Anti-ADAMTS13 autoantibody profiling in patients with immune-mediated thrombotic thrombocytopenic purpura

Abstract

Anti-A Disintegrin and Metalloproteinase with a ThromboSpondin type 1 motif, member 13 (ADAMTS13) autoantibodies cause a severe ADAMTS13 deficiency in immune-mediated thrombotic thrombocytopenic purpura (iTTP). ADAMTS13 consists of a metalloprotease (M), a disintegrin-like (D) domain, 8 thrombospondin type 1 repeats (T1-T8), a cysteine-rich (C), a spacer (S), and 2 CUB domains (CUB1-2). We recently developed a high-throughput epitope mapping assay based on small, nonoverlapping ADAMTS13 fragments (M, DT, CS, T2-T5, T6-T8, CUB1-2). With this assay, we performed a comprehensive epitope mapping using 131 acute-phase samples and for the first time a large group of remission samples (n = 50). Next, samples were stratified according to their immunoprofiles, a field that is largely unexplored in iTTP. Three dominant immunoprofiles were found in acute-phase samples: profile 1: only anti-CS autoantibodies (26.7%); profile 2: both anti-CS and anti-CUB1-2 autoantibodies (12.2%); and profile 3: anti-DT, anti-CS, anti-T2-T5, anti-T6-T8, and anti-CUB1-2 autoantibodies (8.4%). Interestingly, profile 1 was the only dominant immunoprofile in remission samples (52.0%). Clinical data were available for a relatively small number of patients with acute iTTP (>68), and no correlation was found between immunoprofiles and disease severity. Nevertheless, profile 1 was linked with younger and anti-T2-T5 autoantibodies with older age and the absence of anti-CUB1-2 autoantibodies with cerebral involvement. In conclusion, identifying acute phase and remission immunoprofiles in iTTP revealed that anti-CS autoantibodies seem to persist or reappear during remission providing further support for the clinical development of a targeted anti-CS autoantibody therapy. A large cohort study with acute iTTP samples will validate possible links between immunoprofiles or anti-domain autoantibodies and clinical data.

Graphical abstract

Figure 1.

Fine mapping of anti-ADAMTS13 autoantibodies in acute-phase and remission samples. (A) Flow chart representing an overview of the number of patients and their corresponding plasma or serum samples analyzed in this study. A total of 365 plasma or serum samples from 213 patients were selected and screened for the presence of anti-ADAMTS13 autoantibodies. More than 1 sample was available for 92 patients. Samples were both from acute phase (orange) and remission (blue). Positive samples were further screened for the presence or absence of anti-M, anti-DT, anti-CS, anti-T2-T5, anti-T6-T8, and anti-CUB1-2 autoantibodies. From the positive samples, more than 1 sample was available for 27 patients. These samples can be identified in supplemental Table 1 as samples with the same patient identification number but with a different sample identification number. (B) Comparison of the percentage of samples with detectable anti-M, anti-DT, anti-CS, anti-T2-T5, anti-T6-T8, and anti-CUB1-2 autoantibodies between acute-phase samples (n = 131) and remission samples (n = 50). Fisher’s exact test; *P < .05; ***P < .001; ****P < .0001.

Figure 2.

Presence of anti-CS autoantibodies in remission samples with detectable anti-ADAMTS13 autoantibodies according to the ADAMTS13 activity. Difference between groups were assessed by Fisher’s exact test (P < .0001).

Figure 3.

Immunoprofiles in acute-phase (n = 131) and remission (n = 50) samples. The immunoprofiles were formed based on the presence or absence of anti-M, anti-DT, anti-CS, anti-T2-T5, anti-T6-T8, and anti-CUB1-2 autoantibodies. Green rectangles indicate the presence and red rectangles the absence of detectable anti-M, anti-DT, anti-CS, anti-T2-T5, anti-T6-T8, and anti-CUB1-2 autoantibodies. The immunoprofiles are listed starting with the most prevalent immunoprofiles in the acute-phase samples. Orange bars represent the number of acute-phase samples with a specific immunoprofile; blue bars represent the number of remission samples with a specific immunoprofile.

Figure 4.

Link between immunoprofiles or domain-specific anti-ADAMTS13 autoantibodies and Benhamou score, age, and cerebral involvement. (A) Benhamou score according to the 3 most prevalent immunoprofiles. The Benhamou score (<3 [1-2]; ≥3 [3-4]) was calculated based on age, cerebral involvement, and lactate dehydrogenase level. χ² test; P = .4933. (B) Age of patients according to the 3 most prevalent immunoprofiles. Kruskal-Wallis test followed by Dunn’s multiple comparisons test; P = .0333. (C) Age of patients according to the presence and absence of anti-T2-T5 autoantibodies. Mann-Whitney test, P = .0012. (D) Cerebral involvement according to the presence and absence of anti-CUB1-2 autoantibodies. Fisher’s exact test; P = .0177.