Autoimmunity and Novel Therapies in Immune-Mediated Thrombocytopenia

Dana Yehudai,a Elias Toubi,b Yehuda Shoenfeld,c and Zahava Vadaszb

Immune-mediated thrombocytopenic purpura (ITP) is recognized as a cell-specific autoimmune disorder, yet, multifactorial in origin. The development of thrombocytopenia is well proven to be mediated by both humoral (anti-platelet antibodies) and cellular (T-cell) mediated mechanisms. In some cases other autoantibodies are also induced, eg, antinuclear antibody (ANA), anti-dsDNA, and anti-cardiolipin, in addition to anti-platelet antibodies. The persistance of these autoantibodies during the course of ITP could herald future development of another autoimmune disease, eg, systemic lupus erythematosus (SLE) or anti-phospholipid syndrome (APS). Due to the better understanding of the pathophysiology of ITP, new novel therapies were introduced aiming to achieve long-lasting remissions. In this review we will focus on the autoimmune nature of the disease and on some of the mechanisms of action of these new therapies. Semin Hematol 50:S100–S108. C 2013 Published by Elsevier Inc. Open access under CC BY-NC-ND license.

mmune thrombocytopenic purpura (ITP) is con- they require various therapeutic interventions, sidered a classical autoimmune disorder charac- including splenectomy and other novel therapies terized as the occurrence of thrombocytopenia such as rituximab and syk inhibitors.5 Though many I 3 (platelet count o 100 10 /mL. Platelet destruc- reports have pointed to the notion that ITP is tion by the spleen and morphologic abnormalities predominantly a disorder of young women, some of megakaryocytes followed by their altered matura- have shown a progressive increase in incidence with tion are suggested to be the main pathogenic age. However, the demographics of ITP probably abnormalities in ITP. Later, ITP was well defined to depend on regional differences due to different be the result of a plasma factor proven as specific predisposing infections, and the early diagnosis or anti-platelet antibodies and finally defining ITP as an the better treatments of B-cell neoplasms.6 autoimmune disorder.1–3 Recent literature accepts When ITP is isolated and no evidence for systemic the classification of three phases of ITP: (1) the new- diseases is proven, primary ITP is then defined; onset phase that occurs within 3 months from primary ITP represents 80% of all cases. However, diagnosis; (2) the persistent phase, lasting between secondary ITP is defined when various well-proven 3 and 12 months from diagnosis; and (3) the chronic infectious or autoimmune diseases are responsible phase, defined as lasting more than 12 months.4 for the development of thrombocytopenia. These Patients who are diagnosed with ITP who repeti- include viral infections such as hepatitis C virus tively bleed considered to have ‘‘severe’’ ITP and (HCV), cytomegalovirus (CMV), and Helicobacter pylori, and also cases of ITP that were recognized 7,8 aDepartment of Internal Medicine A, Bnai-Zion Medical Center. The as a result of vaccination. Many viral particles such Rappaport Faculty of Medicine, Technion-Haifa, Israel. as HCV were described to polyclonally activate B b Division of Allergy and Clinical Immunology, Bnai-Zion Medical cells by binding CD81 (part of the B cell co-receptor) Center, The Rappaport Faculty of Medicine, Technion-Haifa, Israel. cZabludowicz Center for Autoimmune Diseases, Sheba Medical or other receptors in the case of CMV, predisposing Center, Sakler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, to the expansion of autoreactive B cells, the produc- Israel. tion of anti-viral antibodies that cross-react with Publication of this article was supported by the International Coop- membrane platelet antigens. In some studies the erative ITP Study Group (ICIS). Conflicts of interest: none. presence of multiple autoantibodies was docu- Address correspondence to Yehuda Shoenfeld, MD, FRCP, Zabludowicz mented and suggested to be the result of epitope Center for Autoimmune Diseases, Sheba Medical Center, Tel- spreading.9,10 A wide spectrum of autoimmune dis- Aviv University, Tel-Aviv, Israel. E-mail: [email protected] 0037-1963 eases such as systemic lupus erythematosus (SLE), & 2013 Published by Elsevier Inc. Open access under CC BY-NC-ND license. rheumatoid arthritis (RA), antiphospholipid syn- http://dx.doi.org/10.1053/j.seminhematol.2013.03.015 drome (APS), and immune deficiencies like common

S100 Seminars in Hematology, Vol 50, No 1, Suppl 1, January 2013, pp S100–S108 Autoimmunity and novel therapies S101 variable immune deficiency (CVID) are among the and B cells. The T-cell response was inhibited by list of diseases reported to be responsible for secon- anti-FcgRI antibody. Thus, splenic macrophages that dary ITP. Chronic lymphocytic leukemia (CLL) and take up opsonized platelets via FcgRI are major APCs various lymphomas are additional diseases in the list for cryptic glycoprotein peptides, and are central to of all the above, all of which do not exceed 20%.11–13 the maintenance of anti-platelet autoantibody pro- duction in ITP patients.17 In another study, it was reported that macrophages activated by C-reactive THE DIVERSITY OF IMMUNE-MEDIATED protein (CRP) transfer suppression of ITP. Suppres- ABERRATIONS IN ITP sion of ITP by CRP-treated splenocytes required Early studies showed already that when isolated FcgRI on the donor cell and FcgRIIb in the recipient platelets from ITP patients were incubated with mice. These findings suggest that CRP generates autologous lymphocytes, they induced their trans- suppressive macrophages through FcgRI, which formation and resulted in increased interleukin (IL)-2 then act through an FcgRIIb-dependent pathway in productions. The responsive cells were proven to be the recipient to decrease platelet clearance.18 CD4þ T cells and were shown to react specifically against modified glycoprotein (GP)IIb-IIIa on acti- EFFECTOR VERSUS T-REGULATORY CELLS vated platelets. In addition to CD4þ T-cell activation, it was also noticed that the stimulation of ITP B cells IN ITP resulted in the in vitro production of anti-platelet The immune response to glycoproteins in ITP is antibodies. However, the same stimulation of normal usually modulated by CD4þ effector cells and CD8þ B cells did not induce such autoantibodies. Of cytotoxic cells. In thymus, T cells lose either the interest is that the depletion of CD19þ B cells was CD4 or CD8 antigens and are released as either followed by a complete disappearance of these CD4þ effector or CD8þ cytotoxic T cells (CTLs). antibodies. Using antigen-specific assays such as Some self-reactive T cells survive thymic depletion, immunoprecipitation, immunoblot, and antigen cap- and persist in peripheral blood as autoreactive and ture techniques, these autoantibodies were shown mediate autoimmunity. Normally, autoreactive T to be highly specific, and to recognize antigens that cells are controlled by many peripheral self- are derived from either a single or multiple glyco- tolerance mechanisms such as altered co- proteins. The most identified autoantibodies in ITP stimulatory molecules and or failure of T-regulatory are those against GPIIb-IIIa and/or GPIb-IX, but there cell (Treg) function. When self-tolerance fails, CD4þ have also been reports of autoantibodies against T-effector cells react with specific antigens pre- GPIa-IIa and GPIV.14,15 When assessed positively, sented by MHCII molecules in association with a these autoantibodies are considered a diagnostic proper expression of co-stimulatory molecules. hallmark of ITP with a sensitivity of 49%–66% and When T cells of ITP patients were stimulated by a specificity of 78%–93%; however, negative results platelet antigens, one could notice increased pro- do not rule out this diagnosis.16 The most likely duction of both Th1 and Th2 cytokines. Th1 cells explanation for negative assays is that other mecha- produce IL-2, interferons, and tumor necrosis factor nisms are involved in the pathophysiology of ITP (TNF), whereas Th2 cells produced IL-4, IL-13, and that do not involve anti-platelet antibodies, such as IL-10, which are considered important in inhibiting T-cell–dependent platelet destruction or inhibition cell-mediated immune responses. The role of Th17 of platelet production. (known to be associated with many immune- The persistent destruction of platelets by macro- mediated diseases) in the pathogenesis of chronic phages induces the continuous processing and pre- ITP was assessed. Higher levels of IL-17A and Th17- sentation of the above membrane glycoproteins by related cytokines, and an increased percentage of IL- antigen-presenting cells (APCs), which is a crucial 17A producing CD4þ and neutrophils, were step for the generation of pathogenic anti-platelet observed in ITP patients.19–21 In addition, T- antibodies. Aiming to better understand this process, effector cells are increasingly recruited into bone GPIIb/GPIIIa-reactive T-cell lines generated from ITP marrow and other involved organs in ITP. To patients were cultured with autologous freshly iso- determine the reason for this, expression of chemo- lated splenic macrophages, B cells, or dendritic cells kine receptors such as CX3CRI and CXCR4 was (DCs). Macrophages induced the proliferation of analyzed in the bone marrow of ITP patients. Here, GPIIb/IIIa-reactive T-cell lines without an exogenous T-cell surface expression of these chemokine recep- antigen, but B cells and DCs required glycoprotein tors was increased in patients compared with con- peptides to stimulate the T cells. Cultured macro- trols. Furthermore, the number of CD3þ T cells in phages that captured opsonized platelets promoted bone marrow, but not in blood, along with increased anti-GPIIb/GPIIIa antibody production in mixed cul- Fas expression was also found, emphasizing the tures of autologous GPIIb/GPIIIa-reactive T-cell lines importance of cellular immunity in ITP22 (Figure 1). S102 D. Yehudai et al.

in comparison to healthy individuals. The number of Tregs was significantly lower in ITP patients in the severe phase, and in patients positive for anti-GPIIb/ IIIa antibody. However, the number of those cells increased in patients having full remission, especially after splenectomy. In addition, Foxp3 mRNA levels in peripheral blood mononuclear cells (PBMCs) of ITP patients were higher when patients were in remission than in those with refractory lower plate- let counts.26 In another study, the production and reduced immunosuppressive activity of Tregs in ITP was investigated. The frequency of circulating CD4þCD25þFoxp3þ Tregs in patients and controls was comparable. However, sorted CD4þCD25high cells from patients with chronic ITP (n ¼ 13) had Figure 1. Cellular mechanisms in ITP: the wide rage of a twofold reduction of in vitro immunosuppressive T-cell responses. Following a continuous presentation of GPIIb/IIIa to T cells, Th1 overproduce TNF, IFNs, and IL-6 activity compared with controls (n ¼ 10, P o.05). but also overexpress chemokine receptors such as CXCR4 The impaired suppression was specific to Tregs, as and CXCR1. IL-17 and IL-4 are also overproduced by shown by cross-mixing experiments with T cells TH17 and TH2 cells. Tregs are important in decreasing from controls. These data suggest that functional the proliferation and activation of both TH1 and Th17 in defects in Tregs are involved in the pathogenesis of ITP. Finally, CTLs induce the destruction of mature 27 platelets by overexpressing Fas-ligand and overproducing ITP. The above findings and others have led to the granzymes. understanding that the Treg count can possibly correlate with the severity of ITP and, as such, might Cellular immune-mediated response in ITP is be used as a tool for diagnosis and assessment of frequently manifested by increased proliferation of improvement in ITP patients.28 In spite of their cytotoxic T cells. Many studies alluded to the pivotal role in maintaining peripheral immune toler- capability of cytotoxic T lymphocytes to alter mega- ance, a very small and clinically ineffective number karyocyte development and function, thereby con- of naturally Tregs (nTregs) can be found in periph- tributing to impaired platelet production. In this eral blood. In their work, Zhang et al demonstrated respect, cell-mediated lysis of autologous platelets that platelet GP-specific induced Tregs (GP-iTregs) in chronic ITP was observed using purified CD8þ T could be generated de novo from non-regulatory cells as effector cells. These cells were shown to CD4þCD25CD45RAþ cells in patients with ITP, overexpress Fas-ligand, mRNA levels of granzyme B, and induced both antigen-specific and linked sup- perforin, and TNF-a. These results strengthen the pression. They also showed that the Toll-like recep- concept that apoptosis and perforin/granzyme-medi- tor pathway is the dominant pathway (along with ated cytotoxicity constitute an important pathway notch and transforming growth factor-b pathways) through which CTLs destroy autologous platelets23 related to the GP-specific tolerance.29 (Figure 1). Aiming to assess the beneficial effect of Tregs in preventing ITP in mice, Treg-deficient mice were studied by inoculation of Treg-depleted CD4þCD25 T-REGULATORY CELLS T cells isolated from BALB/c mice into syngeneic The exact mechanisms of autoimmunity in medi- nude mice intravenously. Platelet count, platelet- ating ITP have not been investigated fully, but they associated anti-platelet antibodies, and IgG anti- definitely include a clear imbalance between the platelet antibody production in splenocyte cultures overactivity of effector T cells and the altered were examined by flow cytometry. Of 69 Treg- function of Tregs. Recent advances in the field of deficient mice, 25 (36%) spontaneously developed Tregs established their better molecular definition, thrombocytopenia that lasted at least 5 weeks. The bringing new insights into their role in maintaining platelet-associated IgG level and proportion of reticu- self-tolerance.24,25 Therefore it was obvious to assess lated platelets were elevated in the thrombocyto- the role of these cells in ITP. In patients with ITP, penic mice. The transfer of Tregs efficiently naturally occurring CD4þCD25þ Tregs are both prevented the onset of thrombocytopenia, but Treg functionally impaired and reduced in number. Aim- transfer after the onset of thrombocytopenia had no ing to establish this finding, Tregs were measured in apparent effect. Treatment with IgG anti-cytotoxic T 44 patients with acute ITP. By using flow cytometry lymphocyte-associated antigen 4 (CTLA-4) antibodies analysis, the number of CD4þCD25þ T cells in canceled this Treg-governed suppressive effect. patients with ITP showed a very wide distribution These results indicate that Tregs are capable of Autoimmunity and novel therapies S103 preventing murine autoantibody-mediated thrombo- anti-phosphatidylcholine, anti-phosphatidylserine, cytopenia by engaging CTLA-430 (Figure 1). anti-phosphatidylethanolamine) were more common in ITP than in APS (P o.05); (2) multiple APLA (Z3 antigens) were more frequent in APS than in ITP B-CELL AUTOIMMUNITY IN ITP (P o.05); (3) LA was frequently associated with APS Once tolerance for platelet antigens is lost, few but was absent in ITP; (4) APLA is quite common in pathways can lead to ITP, including antibody-medi- ITP—two-thirds of subjects were positive for at least ated, complement-dependent, and apoptosis. Anti- one APLA. They concluded that the APLA profile platelet autoantibodies can bind to platelets and differs between ITP and APS. In APS, antibodies megakaryocytes, where it has been shown to not were mainly against b2GP1 and 80% had positive only cause platelet destruction but also decrease LA, while in ITP the APLA react mostly with the their production by interfering with megakaryocyte phospholipids without LA. These differences might maturation/proliferation or by causing intramedul- explain the opposite clinical presentations in the lary platelet destruction.31 two diseases: thrombotic versus bleeding. Pratt et al Autoantibody-mediated complement activation on described the prevalence of antithyroid antibodies platelets is a common finding in ITP patients, and its (ATA) and antinuclear antibodies (ANA) in children frequency and specificity have been studied. Najaoui with ITP. Twenty-three percent of children with acute et al32 clearly showed that the major targets for ITP had acute non-platelet antibodies detected. In the complement-fixing autoantibodies are GPIIb/IIIa and chronic ITP patients group, 33% had at least one GPIb/IX, and that in a significant number of patients abnormal antibody value (ATA, ANA). Most of the with chronic ITP, platelet autoantibodies are capable patients testing positive for autoantibodies were of activating the classical complement pathway. female and/or 12 years of age or older. Their results They reported on the presence of complement suggest patients with acute ITP who also have other fixation even in ITP sera with very low titers of autoantibodies may be more likely to develop chronic autoantibodies. One of the potential (of high clinical ITP than those lacking these autoantibodies.37 significance) components of autoimmunity in ITP is BAFF (B-cell activating factor of the TNF family), a crucial cytokine for normal development and sur- THE DEVELOPMENT OF SLE IN ITP PATIENTS vival of B cells33 (Figure 2). In order to investigate this issue, Zhou et al explored the crucial role for The issue of ITP being primary or secondary to BAFF in the pathogenesis of ITP, by measuring BAFF/ SLE was the subject of many studies. When secon- BAFF-R levels and evaluating the connection dary, ITP is usually mild to moderate, and the between clinical parameters and expression levels.34 They found significantly higher serum BAFF levels in untreated ITP patients compared to normal controls and treated ITP patients. A weak correlation between platelet counts and BAFF levels was noticed. However, no statistical difference in BAFF levels between acute and chronic ITP was found. These results questioned BAFF being a marker for ITP disease activity but clearly point to its role in ITP development and autoimmunity. Antiphospholipid antibodies (APLA) are fre- quently found in serum of ITP patients. In their work, Pierrot-Deseilligny et al35 found APLA [anti- cardiolipin (aCL) and/or lupus anticoagulant (LA)] in 26% of 215 ITP patients: anticardiolipin alone in 18%, lupus anticoagulant in 0.5%, and the combi- nation of both in 7%. Thrombotic events, character- istic of anti-phospholipid syndrome (APS), were low (7%), but were associated with age and high IgG-aCL Figure 2. Humoral mechanisms in ITP: anti-platelet levels, suggesting a potential benefit in testing for antibodies: (A). GPIIb/IIIa are presented to Th1 effector aPL when ITP is diagnosed. Bidot et al investigated cells. These enhance the proliferation of autoreactive B the prevalence of APLA in ITP and APS.36 They cells and the production of anti-GPIIb/IIIa antibodies. (B) Autoreactive B cells are stimulated by increased levels observed few differences: (1) IgG and IgM antibodies of BAFF. This is followed by a continuous production of to b2GP1 were more common in APS than ITP, anti-platelet antibodies. (C) Anti-GPIIb/IIIa bind megakar- while IgG antibodies against phospholipids (aCL, yocytes and inhibit their maturation to normal platelets. S104 D. Yehudai et al. treatment in this case is different, especially when positive ANAs. Of the 25 children with positive ITP is present in association with other hematologic ANAs, nine (36%) had further autoimmune symp- findings. Anti-platelet antibodies were reported to be toms during the follow-up period as compared to detected in 78% of SLE patients, often without none in the 62 children with negative ANAs concurrent thrombocytopenia, and in up to 16% of (P o.001). These were predominantly girls (n ¼ 8) patients, isolated thrombocytopenia was their initial with a mean age of 12.2 years at diagnosis and most clinical manifestation. The early recognition of of them had chronic ITP (n ¼ 7) rather than acute. patients in whom isolated ITP is likely to progress Autoimmune symptoms were more frequent in to SLE is crucial due to its both prognostic and those with ANA titers Z1:640 than those with lower therapeutic importance.38–40 titers (67% v 31%), although this difference was not Several studies have attempted to identify clinical statistically significant (P ¼ .20). Moreover, children or laboratory parameters that could possibly predict with a positive ANA together with other autoanti- which of the ITP patients are likely to develop SLE or bodies, such as anti-dsDNA, were significantly more other autoimmune diseases. In this respect, Ander- likely to develop further autoimmune symptoms son and co-authors found that 24 of 117 adult (57% v 0%, P ¼ .04). The author’s conclusion was patients with ITP (20%) had a positive ANA titer. that, similar to the experience noted by many in Four of these patients developed later SLE; all of adults, the presence of a positive ANA (especially of them were women with high titers of ANA.41 In a high titers) together with the presence of other smaller study, six of 18 adult women with ITP (33%) autoantibodies, especially those to dsDNA, can be and positive ANA developed SLE within 4 years, all helpful in identifying those patients with ITP who with high ANA titer (Z1:160).42 In another study, 16 are at increased risk of develop SLE. Several other of 82 adult patients with chronic ITP (20%) had a reports demonstrated later that female patients with positive ANA recorded. Nine of these patients devel- chronic ITP, high titers of ANA, and additional oped SLE within 6 months of ITP onset.43 These data autoantibodies (such as anti-dsDNA, cardiolipin, suggest that high titers of ANA in women with ITP anti-SSB-La/SSA-Ro) are candidates for developing are a sensitive but nonspecific marker for predicting SLE in the future.46–48 Therefore, physicians have the development of SLE. However, one should point to look for a combination of autoantibodies such as to other contrary experiences among children with anti-dsDNA, cardiolipin, and anti-Ro/La, specific chronic ITP and ANA positivity who remained free of medical and familial history, and physical examina- any evidence of autoimmune diseases. With this tion in order to better identify the minority of cases point of view, Kurata et al reported on 29 adult in which the ITP heralds future SLE. patients with chronic ITP and positive ANA, but none of whom developed SLE during a 3-year follow- up period.44 NOVEL THERAPIES FOR ITP In order to further clarify this issue, the clinical The decision for treatment in patients with ITP significance of ANA positivity in children with must take into account several factors, including the chronic ITP was further studied. To do so, a long- age of the patient, the severity of the illness, and the term follow-up study of a large cohort was per- anticipated natural history. Adult patients, particu- formed to determine the likelihood that selected larly those older than 60 years of age, have a higher children with ITP will develop signs and symptoms incidence of major or fatal bleeding than children.49 of systemic autoimmune disease.45 Here, 147 chil- However, therapy may not be necessary unless the dren with ITP were enrolled and divided by sex into platelet count is o20 109/L or when extensive boys (n ¼ 77) and girls (n ¼ 70) and to those defined bleeding develops. This review focuses on recent as having acute (n ¼ 69) or chronic (n ¼ 78) ITP. advances in ITP treatment and not on traditional The average age of these patients was 7.0 Ϯ 4.8 approaches. The main goal of second-line therapy is years. ANA was assessed in only 87 of these patients, to attain a continual increase of the platelet count, typically at the time of diagnosis as a screening test which is considered hemostatic for the patient. The or due to an additional symptoms or a positive family following treatment modalities have quite different history. ANA was found to be positive (a median titer mechanisms of action, but they are intended to of 1:160) in 25 of the 87 tested children. Correla- induce long-term remission. tions between ANA results, family history of auto- immune diseases, initial hemoglobin concentration, initial platelet count, and/or the finding of elevated Rituximab-Specific Anti-CD20 Antibody platelet autoantibodies were not evident during a Rituximab was found to be highly useful in ITP short follow-up. However, when long-term follow-up patients, where 60% of them are usually considered were examined, a higher incidence of autoimmune responders, with 40% achieving complete response. symptoms was noted in the group of children with Responses generally occur between 2–8 weeks of Autoimmunity and novel therapies S105 treatment and last from 2 months in partial respond- rituximab treatment in patients with lymphoma and ers to 5 years or longer in 15%–20% of initially treated and patients with SLE and ITP, who were heavily patients. Initial complete response and prolonged B- immunosuppressed and on combination treatments. cell depletion are considered useful predictors for Thus, additional long-term safety data are required.58,59 sustained responses.50–52 In a prospective study, following one treatment of rituximab, 33% of patients had a platelet count of 50 109/L or higher Thrombopoietin-Receptor Agonists: and 40% had a platelet count of 30 109/L or higher. Romiplostim and Eltrombopag In this respect, most patients with a durable (41 Until now, all known treatments for ITP were year) complete response will respond to a repeated generated in order to suppress the production of treatment when they relapse. In the above studies, autoantibodies and/or inhibiting macrophage- rituximab was used in doses of 375 mg/m2, but lower mediated destruction of opsonized platelets. How- doses (100 mg administered intravenously weekly for ever, some ITP patients have impaired platelet 4 weeks) also may be effective, although associated production rather than increased platelet destruc- with a longer time to response.53,54 Further studies tion. Additionally, ITP patients have normal or are warranted to identify the optimal dose and the slightly elevated thrombopoietin (TPO; the main best treatment protocol. Recently it has been regulator of platelet production) levels. These levels reported that high response rates were shown for a are lower than the serum levels found in patients combination of rituximab with high-dose dexametha- with megakaryocytic hypoplasia, probably as a result sone as initial therapy. At this stage, one should of active TPO uptake and destruction by the question whether rituximab could be a better treat- expanded megakaryocyte mass in ITP. With this in ment choice than splenectomy in order to achieve mind, the uses of TPO for stimulation of megakar- remission as a second-line therapy. Although having yopoiesis and the increase of platelet count in ITP fewer side effects, there are not enough current data were suggested.60 The first-generation of TPO used to support the replacement of splenectomy with in clinical trials included recombinant human TPO rituximab as a second-line treatment for chronic (rhTPO), and a non-glycosylated, truncated form of ITP.55 TPO coupled to polyethylene glycol. The recombi- The main mechanisms by which rituximab is nant protein, called ‘‘megakaryocyte growth and beneficial in patients with ITP is by eliminating differentiation factor’’ (MGDF), had structural and normal B cells, but also those producing anti- immunogenic differences compared to native TPO, platelet antibodies. This B-cell depletion is almost due to which immunogenic adverse effects were always transient and has few side effects or toxicities. noticed.61 Clinical trials were discontinued after the Rituximab was shown also to exert significant effects development of TPO autoantibodies was demon- on cellular immunity. Pretreatment abnormalities of strated in healthy volunteers and did not yield a T cells in ITP patients were reverted in responders of clinically therapeutic advantage.61,62 Recently, how- rituximab as opposed to non-responders, where they ever, intense research and development has been remained unchanged. By being immunomodulatory, focused on second-generation thrombopoietic rituximab produces clinical remission in ITP when growth factors. These new molecules have no expansion of pathogenic T cells is still dependent on structural resemblance to TPO, but still bind and B-cell costimulation. However, when ITP is advanced activate the TPO receptor, and were hence given the and T-lymphocyte clones are more expanded, they name of ‘‘TPO-receptor agonists.’’ Studies have been continue to drive B-cell activity and antibody produc- completed and the US Food and Drug Administration tion, irrespective of the cytokine microenvironment approved two TPO mimetics, romiplostim and produced in turn by these B cells. A large clinical trial eltrombopag. Romiplostim is administered as a 1– supported the notion that at this stage of ITP 10 mg/kg subcutaneous weekly injection. Eltrombo- rituximab is ineffective. In this study, patients with pag is an oral non-peptide TPO-receptor agonist a short duration of disease, in whom presumably T- administered as a 25, 50, or 75 mg daily dose. Data cell expansion was B-cell–dependent, were much from clinical trials have demonstrated that both more likely to respond to treatment than those with drugs are highly effective in increasing the platelet longer ITP duration. Therefore, non-responsiveness count in both healthy volunteers and ITP patients. to anti-CD20 therapy in ITP may be related to an In two placebo-controlled, double-blind randomized inability to modify the oligoclonal nature of the T-cell trials, romiplostim was given to splenectomized and subsets.56,57 non-splenectomized ITP patients for 6 months. An Adverse effects of rituximab are usually mild or overall platelet response rate (44 weeks out of 24 moderate, with a low incidence of infections. There study weeks 450 109/L) was observed in 79% and are also reports of several cases with progressive 88% of romiplostim-treated patients, compared with multifocal leukoencephalopathy associated with 0% and 14% in the respective placebo treatment S106 D. Yehudai et al. group. In these romiplostim studies, it was demon- immune thrombocytopenic purpura by an inhibitor of strated that 87% of romiplostim-treated patients had Syk. Blood. 2009;113:3154–60. reduced or discontinued other modalities of ITP 6. Deane S, Teuber SS, Gershwin ME. The geoepidemiol- therapy, including corticosteroids and intravenous ogy of immune thrombocytopenic purpura. Autoim- mun Rev. 2010;9:A342–9. immunoglobulin. Long-term follow-up of patients 7. Cines DB, Bussel JB, Liebman HA, Luning Park ET. The treated with romiplostim showed that responses ITP syndrome: pathogenic and clinical diversity. were sustained for 4 years on continuous therapy Blood. 2009;113:6511–21. and that most ITP patients were able to decrease or 8. Mantadakis E, Farmaki E, Buchanan GR. Thrombocy- even discontinue concurrent corticosteroid ther- topenic purpura after measles-mumps-rubella vaccina- apy.62,63 This is an important therapy result, since tion: a systematic review of the literature and guidance many of these patients are on immunosuppressive for management. J Pediatr. 2010;156:623–8. treatment for a long period of time. 9. Pileri P, Uematsu Y, campagnoli S, et al. Binding of Similar results have been demonstrated with hepatitis C virus to CD81. Science. 1998;282:938–41. eltrombopag in chronic relapsed or refractory ITP 10. Zhang W, Nardi MA, Borkosky W, Karpatkin S. Role of patients. Although most adverse effects were mild, molecular mimicry of hepatitisC virus protein with the main concern of these therapies is the increased platelet GPIIIa in hepatitis related immune thrombo- cytopenia. Blood. 2009;113:4086–93. bone reticulin, which was found in 10 of 271 11. Bidot CJ, Jy W, Horstman LL, et al. Antiphospholipid patients included in the romiplostim trials and in antibodies in immune thrombocytopenic purpura seven of 117 patients in the eltrombopag trials (o6% tend to emerge in exacerbation and decline in remis- overall). Further studies are warranted in order to sion. Br J Haematol. 2005;128:366–72. determine the importance of this finding. Another 12. Michel M, Chanel V, Galicer L, et al. Autoimmune side effect was liver function test abnormalities, thrombocytopenic purpura and common variable which were seen in 13% of eltrombopag-treated immunodeficiency: analyzes of 21 cases and review patients.63–65 Due to their mechanism of action, of the literature. Medicine (Baltimore). 2004;83: these TPO-receptor agonists are considered as a 254–263. maintenance therapy. Upon cessation of treatment, 13. Hamblin TJ. Autoimmune complications of chronic most patients return to lower platelet counts (410% lymphatic leukemia. Semin Oncol. 2006;33:230–9. 14. Provan D. Characteristics of immune thrombocyto- transiently falling below baseline platelet counts); penic purpura: a guide for clinical practice. Eur J however, few patients are able to discontinue treat- Haematol. 2009;suppl 71:8–12. ment successfully. 15. Chen JF, Yang LH, Chang LX, Feng JJ, Liu Jq. The clinical significance of circulating B cells secreting CONCLUSION anti-glycoprotein IIb/IIIa antibody and platelet glyco- protein IIb/IIIa in patients with primary immune Both humoral and cellular mechanisms are thrombocytopenia. Hematology. 2012;17:283–90. involved in the development of ITP. The role of 16. McMilan R, Wang L, Tani P. Prospective evaluation of Tregs is of special pathogenic importance and could the immunobead assay for the diagnosis of adult be applied in future therapeutic approaches. chronic immune thrombocytopenic purpura. J Patients in whom ANA, aPL, and anti-dsDNA anti- Thromb Haemost. 2003;1:485–91. bodies exist are candidates to develop SLE and or 17. Kuwana M, Okazaki Y, Ikeda Y. Splenic macrophages maintain the anti-platelet autoimmune response via APS. New novel therapeutic approaches are already uptake of opsonized platelets with immune thrombo- in practice for refractory ITP patients. cytopenic purpura. J Thromb Haemost. 2009;7:322–9. 18. Marjon KD, Marnell LL, Mold C, Du Clos TW. Macro- REFERENCES phages activated by C-reactive protein through Fc 1. Cines DB, Blanchette VS. Immune thrombocytopenic gamma RI transfer suppression of immune thrombo- purpura. N Engl J Med. 2002;346:995–1008. cytopenia. J Immunol. 2009;182:1397–403. 2. Cooper N, Bussel J. The pathogenesis of immune 19. Ma D, Zhu X, Zhao P, et al. Profile of Th17 cytokines thrombocytopenic purpura. Br J Haematol. 2006;133: (IL-17, TGF-betta, IL-6) and Th1 cytokines (IFNs) in 364-374. patients with immune thrombocytopenic purpura. 3. McMillan R. The pathogenesis of chronic immune Ann Hematol. 2008;87:899–904. thrombocytopenic purpura. Semin Hematol. 2007;44: 20. Rocha AM, Souza C, Rocha GA, et al. The levels of IL- S3–11. 17A and of the cytokines involved in TH17 cell 4. Rodeghiero F, Stasi R, Gemsheimer T, et al. Stand- commitment are increased in patients with chronic ardization of terminology, definitions and outcome immune thrombocytopenia. Haematologica. 2011;96: criteria in immune thrombocytopenic purpura of 1560–4. adults and children: report from an international 21. Hemdan NY, Birkenmeier G, Wichmann G, Abu El- working group. Blood. 2009;113:2386–93. Saad AM, Krieger T, Conrad K, Sack U. Interleukin-17- 5. Podolanczuk A, Lazarus AH, Crow AR, et al. Of mice producing T helper cells in autoimmunity. Autoimmun and men: an open-label pilot study for treatment of Rev. 2010;9:785–92. Autoimmunity and novel therapies S107

22. Olsson B, Ridell B, Carlsson L, Jacobsson S, Wadenvik 38. Lang BA, Silverman ED. A clinical overview of systemic H. Recruitment of T cells into bone marrow of ITP lupus erythematosus in childhood. Pediatr Rev. patients possibly due to elevated expression of VLA-4 1993;14:194–201. and CX3CR1. Blood. 2008;112:1078–84. 39. Nakamura RM, Bylund DJ. Contemporary concepts for 23. Zhang F, Chu X, Wang L, et al. Cell-mediated lysis of the clinical and laboratory evaluation of systemic autologous platelets in chronic idiopathic thrombocy- lupus erythematosus and ‘‘lupus-like’’ syndrome. topenic purpura. Eur J Haematol. 2006;76:427–31. J Clin Lab Anal. 1994;8:347–59. 24. Miyara M, Gorochov G, Ehrenstein M, Musset L, 40. Mosca M, Tani C, Aringer M, et al. Development of Sakaguchi S, Amoura Z. Human FoxP3þ regulatory T quality indicators to evaluate the monitoring of SLE cells in systemic autoimmune diseases. Autoimm Rev. patients in routine clinical practice. Autoimmun Rev. 2011;10:744–55. 2011;10:383–8. 25. Kornete M, Piccirillo CA. Critical co-stimulatory path- 41. Anderson MJ, Peebles CL, McMillan R, et al. Fluores- ways in the stability of Foxp3þ Treg cell homeostasis cent antinuclear antibodiesand anti-SS-A/Ro in patients with immune thrombocytopenia subsequently devel- in type I diabetes. Autoimmun Rev. 2011;11:104–11. oping systemic lupus erythematosus. Ann Intern Med. 26. Sakakura M, Wada H, Tawara I, et al. Reduced 1985;103:548–50. CD4þCD25þ T cells in patients with idiopathic 42. Perez HD, Katler E, Embury S. Idiopathic thrombocy- thrombocytopenic purpura. Thromb Res. 2007;120: topenic purpura with high-titer, speckled pattern anti- 187–93. nuclear antibodies: possible marker for systemic lupus 27. Yu J, Heck S, Patel V, et al. Defective circulating CD25 erythematosus. Arthritis Rheum. 1985;28:596–7. regulatory T cells in patients with chronic immune 43. Panzer S, Penner E, Graninger W, et al. Antinuclear thrombocytopenic purpura. Blood. 2008;112:1325–8. antibodies in patients with chronic idiopathic auto- 28. Aboul-Fotoh Lel-M MM, El-Deen MA, Osman AM. Role immune thrombocytopenia followed 2-30 years. Am J of CD4þCD25þ T cellsin children with idiopathic Hematol. 1989;32:100–3. thrombocytopenic purpura. J Pediatr Hematol Oncol. 44. Kurata Y, Miyagawa S, Kosugi S, et al. High-titer 2011;33(2):81–5. antinuclear antibodies, anti-SSA/Ro antibodies and 29. Zhang XL, Peng J, Sun JZ, et al. De novo induction of anti-nuclear RNP antibodies in patients with idiopathic platelet-specific CD4(þ)CD25(þ) regulatory T cells thrombocytopenic purpura. Thromb Haemost. from CD4(þ)CD25() cells in patients with idiopathic 1994;71:184–7. thrombocytopenic purpura. Blood. 2009;113: 45. Zimmerman SA, Ware RE. clinical significance of the 2568–77. antinuclear antibody test in selected children with 30. Nishimoto T, Satoh T, Takeuchi T, Ikeda Y, Kuwana idiopathic thrombocytopenic purpura. J Pediatr Hem- M. Critical role of CD4þCD25þ regulatory T cells in atol Oncol. 1997;19:297–303. preventing murine autoantibody-mediated thrombocy- 46. Balsalobre AJ, Herraez HP, Porta EJ, et al. Idiopathic topenia. Exp Hematol. 2012;40:279–89. thrombocytopenic purpura as first manifestation of 31. McMillan R. Autoantibodies and autoantigens in systemic lupus erythematosus. Ann Med Intern. chronic immune thrombocytopenic purpura. Semin 1999;16:611–4. Hematol. 2000;37(3):239–48. 47. Unal I, Ceylan C, Ozdemir F, Turk BG, Kazandi AC, 32. Najaoui A, Bakchoul T, Stoy J, et al. Autoantibody- Kandiloglu G. ITP as an initial manifestation of sub- mediated complement activation on platelets is a acute cutaneous lupus erythematosus. J Dermatol. common finding in patients with immune thrombocy- 2005;32:727–30. topenic purpura. Eur J Haematol. 2012;88(2): 48. Hazzan R, Mukamel M, Yacobovich J, Yaniv I, Tamary 167–174. H. Risk factors for future development of systemic lupus erythematosus in children with idiopathic 33. Youinou P, Pers JO. The late news on BAFF in auto- thrombocytopenic purpura. Pediatr Blood Cancer. immune diseases. Autoimmun Rev. 2010;9:804–6. 2006;47:657–9. 34. Zhou Z, Chen Z, Li H, et al. BAFF and BAFF-R of 49. Provan D, Stasi R, Newland AC, et al. International peripheral blood and spleen mononuclear cells in consensus report on the investigation and manage- idiopathic thrombocytopenic purpura. Autoimmunity. ment of primary immune thrombocytopenia. Blood. 2009;42:112–9. 2010;115:168–86. 35. Pierrot-Deseilligny Despujol C, Michel M, Khellaf M, 50. Arnold DM, Dentali F, Crowther MA, et al. Systematic et al. Antiphospholopid antibodies in adults with review: efficacy and safety of rituximab for adults with immune thrombocytopenic purpura. Br J Haematol. idiopathic thrombocytopenic purpura. Ann Intern 2008;142(4):638–43. Med. 2007;146(1):25–33. 36. Bidot CJ, Jy w, Horstman LL, et al. Antiphospholipid 51. Stasi R, Stipa E, Forte V, Meo P, Amadori S. Variable antibodies (APLA) in immune thrombocytopenic pur- patterns of response to rituximab treatment in adults pura (ITP) and antiphospholipid syndrome (APS). with chronic idiopathic thrombocytopenic purpura. Am J Hematol. 2006;81(6):391–6. Blood. 2002;99:3872–3. 37. Pratt EL, Tarantino MD, Wagner D, et al. Prevalence of 52. Conti F, Perricone C, Ceccarelli F, Valesini G. Ritux- elevated antithyroid antibodies and antinuclear anti- imab treatment of systemic lupus erythematosus in bodies in children with immune thrombocytopenic controlled trials and in clinical practice: two sides of purpura. Am J Hematol. 2005;79(3):175–9. the same coin. Autoimmun Rev. 2010;9:716–20. S108 D. Yehudai et al.

53. Cooper N, Stasi R, Cunningham-Rundles S, et al. The therapy in HIV-negative patients: a report of 57 efficacy and safety of B-cell depletion with anti-CD20 cases from the research on Adverse Drug Events monoclonal antibody in adults with chronic immune and Reports project. Blood. 2009;113(20):4834–40. thrombocytopenic purpura. Br J Haematol. 60. Kuter DJ. New thrombopoitic growth factors. Blood. 2004;125:232–9. 2007;109:4607–16. 54. Patel VL, Mahevas M, Lee SY, et al. Out come at 5 years 61. Kuter DJ, Bussel JB, Lyons RM, et al. Efficacy of following response to Rituximab therapy in children romiplostim in patients with chronic immune throm- and adults with immune thrombocytopenia (ITP). bocytopenic purpura: a double-blind randomized con- Blood. 2012;119:5989–95. trolled trial. Lancet. 2008;371:395–403. 55. Cooper N, Evangelista ML, Amadori S, Stasi R. Should 62. Bussel JB, Kuter DJ, Pullarkat V, et al. Safety and rituximab be used before or after splenectomy in efficacy of long-term treatment with romiplostim in patients with immune thrombocytopenic purpura? thrombocytopenic patients with chronic ITP. Blood. Curr Opin Hematol. 2007;14:642–6. 2009;113(10):2161–71. 56. Stasi R, Evangelista ML, Stipa E, Buccisano F, Venditti 63. Bussel JB, Cheng G, Saleh MN, et al. Eltrombopag for A, Amadori S. Idiopathic thrombocytopenic purpura: the treatment of chronic idiopathic thrombocytopenic current concepts in pathophysiology and manage- ment. Thromb Haemost. 2008;99:4–13. purpura. N Engl J Med. 2007;357:2237–47. 57. Stasi R, Del Poeta G, Stipa E, et al. Response to B-cell 64. Cheng G. Oral Eltrombopag for the long-term treat- depleting therapy with rituximab reverts the abnormal- ment of patients with chronic idiopathic thrombocy- ities of T cell subsets in patients with idiopathic throm- topenic purpura: results of a phase III, double- blind, bocytopenic purpura. Blood. 2007;110(8):2924–30. placebo-controlled study (RAISE) [abstract]. Blood. 58. Fianchi L, Rossi E, Murri R, et al. Severe infectious 2008;112(11):abstr 400. complications in a patient treated with rituximab for 65. Bussel JB, Provan D, Shamsi T, et al. Effect of idiopathic thrombocytopenic purpura. Ann Hematol. eltrombopag on platelet counts and bleeding during 2007;86(3):225–6. treatment of chronic idiopathic thrombocytopenic 59. Carson KR, Evens AM, Richey EA, et al. Progressive purpura: a randomized, double-blind, placebo- multifocal leukoencephalopathy after rituximab controlled trial. Lancet. 2009;373:641–8.