ITP: Tolerance Lost

Despite a seemingly unifying phenotype comprising a reduced platelet count and mucocutaneous bleeding in many newly identified cases, immune thrombocytopenia (ITP) is a disorder of diverse pathogenesis. In this issue of Blood, Cuker and colleagues describe a little-recognized form of secondary ITP that occurred in a minority of patients receiving the anti-CD52 monoclonal antibody alemtuzumab.¹  Their observations may provide insight into a key concept of immunity—tolerance—and supply indirect evidence for previously advanced hypotheses regarding the acquisition of autoimmunity across diverse forms of the disorder.

ITP classically has been defined as arising from autoantibody-mediated accelerated platelet destruction,²  but in recent years additional mechanisms for thrombocytopenia, including cell-mediated autoimmunity³  and insufficient platelet production,⁴  have been identified. In addition, longstanding recognition of distinct clinical presentations has fueled speculation that the molecular and immunologic underpinnings of ITP probably vary considerably among patients. One feature, the rate of spontaneous remission, undoubtedly distinguishes certain patients from others; children with acute ITP experience relapse-free recovery of the platelet count at a strikingly high rate, frequently without any treatment,⁵  whereas new-onset ITP typically becomes chronic in adult patients even if an initial response to immunomodulation has occurred.⁶  Although these observations have provided some clues about potential underlying immunologic aberrancies, the triggers and mechanisms responsible for inciting the autoimmunity have remained largely elusive.

In recognition of these widely diverse natural histories and patterns of responsiveness to treatment, Cines and colleagues have proposed a unifying hypothesis built on the concept of immune tolerance.⁷  Tolerance comprises the mechanisms necessary for ensuring that B and T lymphocytes, which generally are charged with responding to foreign pathogenic invaders, do not recognize self-antigens. By mechanisms including clonal deletion, receptor editing, anergy, activation-induced cell death, and others,^7-10  autoreactive lymphocytes are deleted from the immune repertoire.

According to this model, a defect at a “tolerance checkpoint” during lymphocyte development and maturation may both lead to platelet-directed autoimmunity and define the clinical manifestations of the disorder. Proposed defects in tolerance may occur early in development or during lymphopoiesis in the bone marrow (central defects); in B-cell maturation leading to abnormal repertoires (differentiation); or in mature lymphocytes as a consequence of antigenic stimulation (peripheral defects).⁷  Peripheral defects are posited to be more likely to incur platelet-specific autoimmunity and be durably responsive to immunosuppression or elimination of a stimulatory antigen, whereas central defects are more likely to lead to relapse after therapy, because of a high frequency of autoreactivity among the primary immune repertoire.⁷ 

The observations by Cuker et al may provide further evidence of such a hypothesis. In their report, 2.8% of patients (6 of 216) with multiple sclerosis who were treated with 12 or 24mg/d of alemtuzumab developed ITP a median of 10.5 months after the last exposure to the drug.¹  The disorder was fatal in 1 subject (because of intracranial hemorrhage), severe in 4, and mild in 1. Of the 5 patients who received initial therapy, all responded within a median of 2.5 days. Relapses occurred in 2, requiring additional treatment, but all remained in remission subsequently (median duration of follow-up, 34 months).

This pattern led the authors to postulate a loss of platelet-specific tolerance at a central checkpoint in lymphocyte development, occurring during initial immune reconstitution after administration of alemtuzumab and leading to recognition of platelet self-antigens. The autoreactivity dissipated presumably through the deletion of the self-reactive lymphocyte clone(s) during subsequent expansion of the lymphocyte repertoire (see figure). The significant delay in the onset of thrombocytopenia after last exposure to the drug would appear to be consistent with their hypothesis, as would the rapid responsiveness to first-line therapy and the long-lasting remissions. Cuker and colleagues distinguished this clinical pattern from Evans syndrome, which typically has little potential to remit durably after initial immunomodulation,¹¹  presumably because of an enduring defect in a central tolerance checkpoint. In contrast, the clinical course of the alemtuzumab-treated patients is reminiscent of the usual pattern of spontaneous remission in acute childhood ITP, although in that disorder a peripheral perturbation in tolerance because of transient stimulation by an antigen is more likely.⁵ 

Although tantalizing, this hypothesis is limited. If a singular mechanism (a defect in a central tolerance checkpoint allowing for growth of an autoreactive lymphocyte clone) was causative, why were there any differences in the clinical manifestations among the patients? Why was the severity of thrombocytopenia not uniform? Why, if further immune recovery led to deletion of the autoreactive species, did thrombocytopenia persist in 2 of the patients, requiring additional immunomodulation? Why were platelet-bound antibodies detectable in only 3 of 5 patients, and why was their appearance and disappearance coincident with the onset and resolution of thrombocytopenia in even fewer patients? Even in this well-defined group of individuals, whose ITP almost certainly was because of a shared pathogenic immunologic mechanism, undeniable heterogeneity suggests the presence of additional disease-modifying factors.

Nonetheless, some conclusions are possible. Alemtuzumab-associated ITP is highly immunoresponsive and has a favorable prognosis, which may direct management away from splenectomy, even though severe thrombocytopenia is expected and relapses after first-line treatment may occur. Implications of these observations for commoner presentations of ITP, however, are less certain. Truly individualized treatment of adult primary ITP, for instance, which exhibits widely ranging variances in clinical features at presentation, in responsiveness to treatment, and in rates of spontaneous remission,⁶  continues to be elusive. Clinical heterogeneity almost certainly indicates inter-individual differences in response to a shared pathogenic mechanism, different pathogenic mechanisms, or both.

Improved outcomes encompassing increased numbers of durable remissions and reduced untoward iatrogenic effects are within our grasp as the understanding of the complex role of the immune system—and a failure to maintain tolerance of self-antigens—continues to emerge.