Relapse and iTTP: beyond the ADAMTS13 activity

In this issue of Blood, Prasannan et al report very interesting work that studied the relationship of either an open or closed ADAMTS13 protease conformation at the time of peak ADAMTS13 activity, noting an eightfold decrease in the relapse risk in the first year and a fivefold decrease in the risk of relapse at 2 years in patients with a closed or folded conformation vs an open ADAMTS13 conformation.¹ We are seeing dramatic changes and improvements in the upfront treatment of immune-mediated thrombotic thrombocytopenic purpura (iTTP), including the use of targeted therapies including caplacizumab, which has reduced the rates of refractory iTTP and early recurrences or exacerbations of iTTP.² As such, increasing attention is now being focused on understanding the risk for relapse of iTTP and the potential factors in addition to the ADAMTS13 activity to refine our ability to predict who may be at increased risk for relapse and who may benefit from more intense immune modulatory therapy.

Presently the in vitro ADAMTS13 activity alone (ADAMTS13 relapse³) is used to judge the risk of relapse and the indication for preemptive rituximab in remission to prevent relapse. This is based in part on work from Jin et al that demonstrated that the risk of relapse in the 3 months following the remission ADAMTS13 activity measurement was increased with decreasing ADAMTS13 activity.⁴ These data were also the first clue that other (non-ADAMTS13) factors may contribute to the risk as younger patients were shown to be at greater relapse risk than older patients given similar ADAMTS13 activity. Interesting work recently from Roose et al has also demonstrated that an open ADAMTS13 conformation is a hallmark of acute iTTP, but it can also be seen in patients in remission and with ADAMTS13 activity greater than 50%.⁵ De Waele et al⁶ then showed that immunoglobulin G antibodies targeting the cysteine/spacer domains of ADAMTS13 were able to induce an open conformation, suggesting a potential role for autoantibodies binding these epitopes, leading to either a de novo or relapsed iTTP episode. These findings have led to additional work including this article from Prasannan et al to study the ADAMTS13 conformation in remission as a predictor of the risk of relapse of iTTP even in those with normal ADAMTS13 activity.

These data from Prasannan et al are a significant step forward in optimizing our ability to predict relapse beyond the use of the ADAMTS13 activity alone. Severely deficient ADAMTS13 activity alone is not sufficient for an acute TTP episode to develop as demonstrated in mouse models of TTP⁷ and congenital TTP patients that present in early adulthood at the time of their first pregnancy.⁸ A “second hit” in the form of pregnancy, infection, or other inflammatory events is required to trigger an acute episode of both iTTP as well as congenital TTP. It would make sense then that risk factors in addition to the ADAMTS13 may improve the prediction of clinical relapse. Prasannan et al reported a change to the open conformation (and an increased relapse risk) that preceded the detection of an anti-ADAMTS13 antibodies. This may indicate that the ADAMT13 conformation change may occur as a consequence of anti-ADAMTS13 antibodies before they increase to the level of detection or may potentially be evidence to suggest another mechanism for the change in the ADAMTS13 conformation to the greater-risk open conformation. Additional non-ADAMTS13 factors have also been studied and may also factor into the risk of relapse including the ratio of high-molecular-weight multimers to low-molecular-weight multimers in remission,⁹ complement activation, and the presence of circulating ultralarge von Willebrand factor (VWF) multimers, which may not always correlate with the ADAMTS13 activity.¹⁰ These biomarkers are also be affected by clinical factors including pregnancy, inflammation, surgery, and vaccinations, creating a model of multiple interacting endogenous and acquired risk factors that could be viewed as similar to how we evaluate the risk for recurrent venous thrombosis.

Ultimately, risk prediction modeling that incorporates multiple variables may be required to predict those at greatest risk for relapse (see figure). The risk of relapse in remission is not likely constant, but more likely fluctuates with the interaction of multiple factors in the context of the ADAMTS13 activity. This would further argue for regular, longitudinal measurement of the ADAMTS13 activity and other informative biomarkers to estimate the relapse risk and need for immune suppressive therapy.

Conflict-of-interest disclosure: The author declares no competing financial interest.