ADAMTS13 and thrombotic thrombocytopenic purpura: where we are and where we need to go

The discovery of the important role of the von Willebrand factor (VWF)-cleaving protease (ADAMTS13) in the etiology and pathogenesis of thrombotic thrombocytopenic purpura (TTP) has enlivened the study of this rare disorder. In the 15 years since the discovery of ADAMTS13, publications on TTP have more than doubled compared to the seven decades before its discovery. It is now rare to encounter a scholarly work on TTP without mention of ADAMTS13. With its critical role in modulating VWF activity, the clotting protein primarily responsible for initial platelet adhesion, ADAMTS13 has been measured enzymatically and antigenically, subjected to numerous structure function analyses, knocked out, patented, expressed in recombinant form, and infused into animals and humans and has been called by many people the cause of TTP. Its activity and antigenic levels, and levels of autoantibodies against it, have been studied in many hundreds of patients with TTP and other diseases. Yet, for all this analysis, the clinical utility of measuring ADAMTS13 in TTP patients remains to be established.

In this edition of TRANSFUSION, Bentley and colleagues¹ have analyzed clinical and laboratory data on 110 patients with clinically suspected TTP with the goal of identifying readily available clinical factors that predict low ADAMTS13 levels. The authors devised an algorithm by which clinicians can estimate ADAMTS13 activity levels in suspected TTP patients without having to measure it. The study is based on the premise that, in that absence of a readily available clinical assay, a surrogate method to estimate the functional level of ADAMTS13 could provide important guidance for initial patient management.

As we consider the value of a surrogate method of estimating ADAMTS13 activity, it may be wise to take stock of where we are in the understanding of ADAMTS13 and TTP. Although it may be tempting to declare that we already understand enough about ADAMTS13 to make it an integral part of TTP patient care, in reality, there is very challenging work that remains before all the information we have about ADAMTS13 can be put to optimal use.

Two critical factors are missing in the understanding of the relationship between ADAMTS13 and TTP, and their absence prevents our knowledge from reaching its full potential: 1) There is no consensus definition of TTP and 2) there is no consensus definition of ADAMTS13 deficiency. These are challenging and potentially contentious issues, but until standardized definitions for TTP and ADAMTS13 deficiency are developed, clinicians, investigators, and authors will be left talking past each other—potentially at the expense of lost opportunities to make major strides in the field.

To clearly understand the importance of a factor like ADAMTS13 in a disease like TTP, a key requirement is to have a standardized definition of the disease. If a disease can be clearly defined in an agreed-upon manner, then diagnostic markers, treatments, or clinical features can be studied by different investigators in cohorts of specifically defined patients. The value of this approach is that results of studies can be tested for reproducibility by other investigators, using cohorts of similar patients, and the results can be extrapolated to patients who meet the defining criteria. Without a standardized definition of disease, cohorts of patients will differ, and results cannot be directly compared between studies.

There has never been a standardized definition of TTP. For decades, the terms TTP and hemolytic uremic syndrome (HUS) have been used without clear distinction, often as HUS/TTP, because disease-defining criteria do not exist to precisely distinguish one from the other.^2,3 Therefore, for all the talk of TTP, we still cannot say exactly what it is. A number of studies define a classic pentad of clinical findings (fever, encephalopathy, microangiopathic hemolysis, renal failure, and thrombocytopenia), but this pentad of findings has never been adopted as the diagnostic definition of TTP (and studies rarely if ever cite these five criteria as their case definitions). Most authors agree that all TTP patients have thrombocytopenia and microangiopathic hemolytic anemia, but those two features are shared with many other disorders. In addition, the classic pentad of TTP findings is seen in disorders other than TTP, such as disseminated intravascular coagulation.

In the absence of a standardized definition of TTP, authors and clinicians have applied their own definitions. Without a precise answer to the question of what is TTP, however, we cannot have a precise way of studying and understanding ADAMTS13 levels in TTP.

Although some authors have suggested that severe deficiency of ADAMTS13 in the setting of other features of TTP defines TTP,^4,5 severe ADAMTS13 deficiency has not been established as a diagnostic criterion for TTP. Scholars have reported markedly different rates of ADAMTS13 deficiency in TTP, ranging from 16% to 100%.^4,6-11 How can we reconcile these widely discrepant figures without acknowledging that a key difference is the disease-defining criteria that were used in making the diagnosis of TTP?

It may be that ADAMTS13 deficiency (and inhibitory antibodies) should be a part of the definition of TTP. To make that connection we also need a definition of ADAMTS13 deficiency. The different methods for assaying ADAMTS13 activity and different definitions of severe deficiency used by investigators create another challenge. If we are to adopt ADAMTS13 deficiency as a defining criterion of TTP, we must first decide what it means. In clinical studies, authors have set the criterion for severe ADAMTS13 deficiency anywhere between less than 5 and less than 15% of normal plasma activity. Reported data on ADAMTS13 activity in thousands of subjects suggest that establishing a cutoff we can all agree upon may be challenging.¹² There appears to be a continuum of activity levels ranging from less than 6% to more than 100%, with hundreds of subjects in every decile of activity level.¹² So, what levels constitute severe deficiency of ADAMTS13 activity?

The most important purpose for defining TTP and ADAMTS13 deficiency is to guide treatment. At this time, for all the foregoing reasons, using ADAMTS13 measurements to guide therapy, especially to decide whether or not to treat with plasma exchange, cannot be justified. It is one thing to risk imposing plasma exchange on a patient that—if we only knew—did not need it. But, to withhold plasma exchange treatment in a patient with suspected TTP on the grounds that a measurement of ADAMTS13 indicates they do not have TTP or will not respond to plasma exchange, is a decision that is not supported by current published data.¹³

Before we progress much further in the study of TTP and ADAMTS13, it is wise to give serious consideration to developing standardized definitions for TTP. The value of such an exercise is illustrated by a number of examples: For many decades, the World Health Organization has periodically convened multinational committees to develop and refine tumor classification criteria. These criteria allow the identification of well-defined cohorts of patients that are critical to clinical trial design and the advancement of cancer treatments. A consensus approach was also taken with the previously poorly defined entity transfusion-related acute lung injury (TRALI).¹⁴ Consensus definitions now allow critical assessment of matters relating to TRALI. Nothing similar has been done with TTP, and it is long overdue.

Creating consensus definitions for TTP would be a challenge, but not an insurmountable task. For one thing, there does not need to be a single definition of TTP. It may be reasonable to borrow from a model such as von Willebrand disease, which has been parsed by expert panels into a number of defined subtypes. For example, there could be Type 1 TTP, defined by certain clinical criteria plus ADAMTS13 deficiency below a certain level (using standardized tests) or, perhaps, Type 1A defined by a genetic deficiency of ADAMTS13 and a Type 1B defined by inhibitor-mediated deficiency. TTP with different clinical features and different levels of ADAMTS13 activity could be defined. There are many possible schemes that could be adopted with emphasis on consistency and practicality.

As a consensus panel wrestles with the question of how to incorporate ADAMTS13 activity and inhibitors into TTP disease classification, the abundance of data accumulating in reference laboratories on ADAMTS13 activity and inhibitor levels may provide a framework for making these decisions.¹² Importantly, the process should remain pragmatic, with a goal of consensus decisions to facilitate clinical studies. Therapeutic recommendations should not be promulgated by a consensus committee, but must be determined in clinical studies using consensus-derived disease-defining criteria.

ADAMTS13 levels are not without clinical utility today. Several studies have shown that TTP patients with ADAMTS13 deficiency are far more prone to relapse than patients without ADAMTS13 deficiency, and their time to respond to plasma exchange is longer.^6,13 Recent data suggest that higher levels of ADAMTS13 inhibitor are a significant risk factor for TTP-related mortality. Such information may reasonably influence clinical studies or individual patient care plans, if used judiciously. But major decisions, such as whether or how to treat patients with suspected TTP, cannot justifiably be based on ADAMTS13 levels at this time. Perhaps nothing demonstrates this conundrum better than the design of the STAR (Study of Rituximab and TTP) clinical trial by the US Transfusion Medicine Hemostasis Clinical Trials Network. The STAR protocol committee, composed of leading experts in the field of TTP, did not see fit to use ADAMTS13 studies either in the disease-defining criteria for the trial or for stratifying patients into treatment groups.

The yearning for further enlightenment on TTP continues to motivate investigators to study ADAMTS13 and to seek meaning in its values for patient care. With each new effort another tile in the mosaic is laid. If only we had a consensus definition of TTP we might be able to determine how ADAMTS13 activity, antigen, or antibody levels are informative in identifying and treating this disease. TTP begs for a consensus definition, and all of the hard work in characterizing ADAMTS13 and other factors in TTP will fall short of its potential until that definition becomes a reality.

CONFLICT OF INTEREST

The author declares no conflict of interest.