Frontiers in Clinical Drug Research - Hematology: Volume 3

By Bentham Science Publishers

Frontiers in Clinical Drug Research – Hematology is a book series that brings updated reviews to readers interested in learning about advances in the development of pharmaceutical agents for the treatment of hematological disorders. The scope of the book

• Language: English
• Publisher: Bentham Science Publishers
• Release date: Apr 5, 2018
• ISBN: 9781681083674

Book preview

Advances in the Understanding and Treatment of Immune Thrombocytopenia

José Perdomo*

Haematology Research Unit, St George and Sutherland Clinical School, University of New South Wales, Australia.

Abstract

Immune thrombocytopenia (ITP) is an acquired autoimmune disorder characterized by a platelet count of less than 100 x 10⁹ platelets/L. ITP results from two distinct processes: accelerated platelet destruction and reduced platelet production. A distinction of the relative contribution of these pathologies should help guide more targeted treatment decisions. Mechanistically, decreased platelet production is caused by autoantibody-mediated damage to megakaryocytes, while increased clearance of antibody opsonized platelets has traditionally been attributed to the activity of splenic and hepatic macrophages. T cell mediated toxicity has also been described as a contributor to ITP pathogenesis. Recent observations of increased platelet apoptosis and glycoprotein desialylation associated with platelet clearance by hepatocytes provide new avenues for therapeutic intervention. The aim of ITP therapy is to attain sufficient platelet levels to achieve haemostasis. Significant improvements have been obtained with first line therapies such as corticosteroids and intravenous immunoglobulins. For unresponsive patients, second line therapies (splenectomy, rituximab, TPO receptor agonists) have proved beneficial. Nevertheless, the heterogeneous nature of ITP demands further understanding of the causal biological processes to provide personalized and more effective therapies. This chapter presents an account of the current understanding of the biology of ITP and discusses the existing and potential new treatments.

Keywords: Apoptosis, Autoantibodies, Autoimmunity, Desialylation, Immune thrombocytopenia, Immune thrombocytopenic purpura, ITP, IVIg, Megakaryocyte, Platelets, Splenectomy, Thrombopoiesis, TPO receptor agonists.

---

* Corresponding author José Perdomo: Haematology Research Unit, St George and Sutherland Clinical School, University of New South Wales, Australia; Tel: [phone]; Fax: [phone]; E-mail: [e-mail]

INTRODUCTION

Thrombocytopenia (low platelet count) is caused by several conditions of both immune or non-immune nature. The term immune thrombocytopenia (ITP) (previously known as idiopathic thrombocytopenic purpura) refers to an acquired

condition of autoimmune origin. Thrombocytopenia due to other causes - such as viral infections, myelosuppressive therapy, bone marrow disorders, Helicobacter pylori infection and drugs - is termed secondary ITP. Unlike secondary ITP, primary ITP is of unknown aetiology.

This Chapter will concentrate on primary ITP, will discuss current treatment strategies and will highlight the latest advances in the understanding of ITP biology and the potential for diagnosis and novel therapies.

A condition consistent with symptoms of ITP was termed Werlhof’s disease and was described in the 18th century [1]. After the discovery of platelets and their role in haemostasis it was understood that the disorder then known as purpura was due to reduced platelet numbers. The incidence of ITP is up to 3.3 cases per 100, 000 people per annum [2]. Prevalence ranges from 9.5 to 23.6 cases per 100, 000 people [3, 4]. The higher prevalence figure is due to the chronic nature of ITP in many patients.

Platelet Biogenesis

Platelets are enucleated, discoid cells, 2-3 µm in diameter with a normal range in healthy adults of 150 – 400 x 10⁹ per L of peripheral blood. Platelets are derived from megakaryocytes (MK), the platelet progenitor cells, which in turn are derived from haematopoietic stem cells. A number of specific transcription factors and cytokines drive megakaryocytic differentiation. Absence of the leucine zipper factor NF-E2 results in severe thrombocytopenia [5] while forced expression of NF-E2 boosts MK differentiation and platelet release [6]. NF-E2 enhances the expression of 3β-hydroxysteroid dehydrogenase, which leads to augmented estradiol production within MK. This autocrine hormonal signaling then triggers proplatelet formation (proplatelets are cytoplasmic branched projections from which platelets are released) [7]. Estrogen treatment increases MK population in the human bone marrow [8] and high estrogen doses induce MK differentiation and a corresponding increase in platelet formation in mice [9]. The role of the female hormone in thrombopoiesis is likely to be behind the observation of higher platelet counts in female adolescents [10], which correlates with increased estrogen production levels to maintain sexual characteristics. NF-E2 appears to be critical for the final stages of MK maturation and for proplatelet formation. This is exemplified by the capacity of NF-E2-/- cells to become MK which then fail to produce proplatelets [11].

The zinc finger proteins GATA-1 and its transcriptional co-regulator FOG-1 are essential haematopoietic factors and their absence causes embryonic lethality. Specific disruption of GATA-1 in the megakaryocytic lineage results in accumulation of immature MK and thrombocytopenia, while severe MK defects are observed in mice lacking FOG-1. Cells lacking GATA-1 fail to differentiate in vitro and in the presence of thrombopoietin, proliferate indefinitely in an immature state [12]. Importantly, restoration of GATA-1 expression results in terminal maturation into megakaryocytic and erythroid lineages [12]. Embryonic stem cells where GATA-1 was silenced by an inducible shRNA construct also proliferate in vitro in an undifferentiated stage. Upon resumption of endogenous GATA-1 expression these cells undergo differentiation and are capable of producing functional platelets in vivo [13].

The importance of the GATA-1/FOG-1 axis in platelet biology is exemplified by a point mutation in the glycoprotein Ib beta promoter that gives rise to a type of the Bernard-Soulier syndrome [14]. In addition, Ets domain transcription factors, specifically Fli-1, are thought to act together with GATA-1/FOG-1 in MK development. Absence of Fli-1 disrupts MK development and hemizygous expression of Fli-1 has been founds in patients with the Jacobsen or Paris-Trousseau Syndrome [15].

During differentiation from a myeloid progenitor, megakaryocytic cells undergo sequential changes that include an initial stage of proliferation followed by differentiation and endomytosis without cell division. This gives rise to large, polyploid (4N, 8N, 16N, 32N, 64N) MK that maintain expression of the amplified genes. This magnified gene expression is likely to contribute to additional protein accumulation within the enlarged cell [16] and to a buildup of platelet-specific proteins.

Apart from the transcriptional regulators mentioned above, MK differentiation and platelet release depend on the activity of cytokines in the microenvironment, in particular thrombopoietin (TPO), stem cell factor (SCF), interleukin (IL) 3 and IL-11. TPO is the ligand for the c-Mpl receptor (CD110) and is considered the principal cytokine that stimulates megakaryopoiesis [17]. TPO is required throughout MK development. Binding of TPO to c-Mpl causes receptor dimerization, activation of Jak2 followed by Jak2 self-phosphorylation and phosphorylation of c-Mpl. These phosphorylation steps are the triggers for intracellular signaling through the STAT and PI3K pathways.

The consequences of changes in c-Mpl function are illustrated by observation of thrombocytopenia in patients with mutations causing c-Mpl loss or reduced function and thrombocytosis in those with constitutively activating mutations (reviewed by [18]). In mice, absence of TPO or c-Mpl causes a marked decrease (around 90%) in circulating platelet numbers. Of note, mice lacking both TPO and c-Mpl do no show further drop in platelet numbers. Moreover, the remaining platelets in these animals are functional in several assays and mice do no suffer from haemorrhagic complications [19]. These observations indicate that to some extent other factors can drive MK differentiation and production of normal platelets. In this sense, the primary role of TPO appears directed towards control of platelet numbers. It is important to note that TPO and c-Mpl are also centrally involved in the maintenance and expansion [20] and in the repair of DNA breaks of adult haematopoietic stem cells [21].

Evidence that other cytokines such as IL-3, IL-6 and IL-11 possessed thrombopoietic function emerged in the 1990s. It was shown that IL-11 alone or together with IL-3 [22] or IL-11 in combination with IL-6 promoted MK differentiation in vitro [23]. However, the hypothesis that these cytokines compensate for the absence of TPO or are responsible for certain level of MK differentiation or platelet production was not supported by in vivo observations. For example, mice deficient in c-Mpl and IL-3 receptor alpha did not show further reduction in platelet or MK numbers [24], indicating that in vivo IL-3 activity did not contribute to MK differentiation. Moreover, mice lacking both c-Mpl and either IL-6 or IL-11 receptor alpha did not show further reduction in platelet counts [25]. Indeed, platelet numbers in mice lacking either IL-6 or IL-11 receptors were no different from wild type [25]. It was later shown that the stimulatory effect of IL-6 observed in inflammatory thrombocytosis is actually mediated by TPO [26]. In vitro, IL-11 has been observed to stimulate MK maturation but, unlike TPO, does not promote proliferation [27]. In cases of cancer patients undergoing myeloablative therapy an inverse correlation between IL-11 levels and platelet counts has been observed [28]. In addition, administration of recombinant IL-11 increases MK and platelet counts in chemotherapy-related thrombocytopenia [29], indicating that in these conditions IL-11 is involved in thrombopoiesis. Altogether, these findings indicate that in normal physiology IL-3, IL-6 and IL-11 do not contribute in a significant manner to MK differentiation and platelet production. Stromal cell-derived factor-1 (SDF-1, also known as CXCL12) is involved in MK localization within the bone marrow. Exogenous SDF-1 causes MK re-distribution and increases platelet production. Even though SDF-1 does not affect platelet production in vitro, its activity in the bone marrow indicates that it is important for thrombopoiesis [30].

An unexpected finding is that TPO appears to be dispensable for the final stages of proplatelet formation and platelet release [31]. In fact, there are indications that TPO has an inhibitory effect on proplatelet formation in vitro [32, 33]; however this inhibitory effect may in fact reflect additional maturation (e.g. endomytosis) of MK in the presence of TPO which causes a delay in pro-platelet formation. How this may play out in vivo is yet to be determined.

Platelet Function

The function of platelets as the principal agents of haemostasis has been recognised for a long time. The pioneering work of Bizzozero in the 1880s reached the conclusion ‘La conclusione delle mie ricerche non possa essere altro che questa: che la parte principale nella coagulazione del sangue spetta non ai globuli bianchi ma alle piastrine’ [34] ‘the conclusion of my investigations could not be other than this: that the principal part for blood coagulation resides not in the white cells but in the platelets’ (author’s translation). These observations, together with later findings regarding the role of platelets in the regulation of vascular permeability, cemented the role of platelets as the keepers of vascular integrity.

Despite their well-established function in haemostasis, platelets have also critical roles in both innate and acquired immunity, bacterial and viral infections and inflammatory disease. That platelets have additional roles seems clear in retrospect, given that the normal levels of circulating platelets are more than sufficient for what is required for haemostasis. For instance, platelet counts of 100 x 10⁹/L are sufficient to perform surgery [35].

Platelets are involved in acute phase immune response to infection [36]. In this case, their main role may be platelet activation and clotting to ensnare the invading organism and restrict the infection. Platelets express Toll-like receptors (TLR) [37] which are well known for their function in infection. TLRs interact with molecules such as lipopolysaccharide (LPS) from Gram negative bacteria and double stranded RNA. Interactions with LPS cause platelet activation and microparticle (vesicles of 0.1 to 1 µm) release which leads to thrombocytopenia and platelet sequestration in the lungs. Roles for platelets in acquired immunity, regulation of gene expression via micro RNA and regulation of vascular inflammation have also been documented [38].

One relatively recent finding is the active secretion of DNA and other cellular proteins such as histones, myeloperoxidase and elastase by activated neutrophils. In this process known as NETosis, neutrophil extracellular traps (NETs) are formed to entangle invading organisms and constrain their capacity to spread [39]. Leukocyte-platelet interactions have been known for many years and are mediated by P-selectin and its receptor on leukocytes, P-selectin glycoprotein ligand-1 (PSGL-1) [40]. There are also interactions between GPIb on platelets and leukocyte CD11b/CD18 [41]. Neutrophils can also interact with platelets via fibrinogen bound to GPIIb/IIIa [42]. In sepsis, activated platelets interact with neutrophils and induce neutrophil activation and induction of NETs [43]. Since a key role of NETosis is fighting infection, and platelets are central components of both NETosis induction and of NETs structure, platelets can also be considered innate immune cells [38].

Epidemiology, Aetiology and Disease Progression

Primary ITP is defined as an isolated platelet counts of < 100 x 10⁹/L (reference count 150 – 400 x 10⁹/L) in the absence of other causes or conditions that may cause thrombocytopenia [44]. The presence of bleeding is not a diagnostic criterion for ITP. ITP is a relatively uncommon disease and its annual incidence has been estimated at 3.3 per 100,000 [2] (this assessment is based on European studies and there could be variations elsewhere). The prevalence of ITP has been calculated at 9.5 per 100,000 [4]. ITP is a condition of both children and adults, but paediatric ITP has some unique characteristics and tends to resolve spontaneously [45]. The discussion here refers to adult ITP only.

Population studies conducted after 2000 have questioned earlier ideas that ITP affected predominantly young women. The median age for ITP patients is 56, but the frequency of ITP increases with age, with an incidence of more than double for those over 60 [46]. This might be due to changes in the immune system and susceptibility to infections in older patients. For younger patients (those under 60) there is a somewhat higher prevalence of ITP in women (ratio 1.7), but in the older cohort the male/female ratio approaches 1 [46]. There is diversity in presentation, with some patients showing signs of bleeding (petechiae, mucosal bleeding), while other patients have less bleeding tendencies even when presenting with very low platelet numbers [47].

The aetiology of ITP is yet to be fully described, nevertheless the basis for low platelet counts falls within the general categories of increased or accelerated platelet destruction and reduced platelet production. Due to its causative diversity and presentation it is now clear that ITP is not a discreet disease and this heterogeneity has led to some authors proposing the term ITP syndrome [47]. A plethora of triggers has been described for secondary ITP. These include Helicobacter pylori infection, viral infections (HIV, hepatitis C and cytomegalovirus), anti phospholipid syndrome, systemic lupus erythematosus, Evans syndrome, malignancy, some vaccines and transplantation.

Several drugs such as quinine, quinidine, rifampicin and vancomycin also cause ITP (for a recent review see Chong et al. [48]). Heparin, which can cause heparin- induced thrombocytopenia, is considered a distinct type of reaction and is not included in ITP-like conditions. In addition, there are several non-immune conditions that cause thrombocytopenia that should also be considered, for instance disseminated intravascular coagulation, thrombotic thrombocytopenic purpura, bone marrow disorders, liver disease, congenital thrombocytopenia, chemo- and radio- therapy and portal hypertension. Primary ITP, on the other hand, is a diagnosis of exclusion and by definition there is no particular causative agent or specific laboratory tests. It is highly probable that many cases diagnosed as primary ITP are in fact secondary ITP with the cause yet to be identified. Currently there are diverse mechanisms implicated in the development and progression of primary ITP.

Platelet Destruction in ITP

The demonstration in the 1950s by the well-documented Harrington-Hollingworth experiment [49] demonstrated that a factor in blood was responsible for platelet destruction. In this setting no obvious changes were found in the MK of test subjects after examination of bone marrow aspirates, indicating that the effect was due to platelet destruction in the peripheral blood. These investigators also examined one patient in more detail and showed that the responsible factor was present in the globulin fraction of plasma [49]. Subsequently, it was revealed that these activities were IgG autoantibodies that recognised platelet antigens [50-52].

The nature of these antiplatelet autoantibodies is consistent with expansion of antibody expressing B cell clones, suggesting that the development of platelet-specific antibodies in ITP is driven by persistent stimuli by platelet antigens [53]. Platelet reactive IgA and IgM antibodies are less common but they are also found in ITP patients [54, 55]. An important consideration is the fact that autoantibodies are not detected in all ITP patients. Different proportions (ranging from 20% to almost 80%) have been reported [50-52, 55, 56]. The variations observed are likely due to the heterogeneous nature of ITP, the number of patients analysed in different studies and the methodologies used for autoantibody detection. Overall 30-40% of patients have no detectable autoantibodies [57].

Autoantibodies recognise abundant platelet glycoproteins, mainly the fibrinogen receptor, GPIIb/IIIa or the von Willebrand factor receptor, GPIb/IX [58, 59]. Antibodies against other platelet antigens such as GPIV and GPIa/IIa have also been described [54, 55, 60] but appear to be rare. It is likely that only autoantibodies against abundant platelet glycoproteins manifest clinical ITP signs and are therefore found more often in laboratory analyses. Opsonized platelets are destroyed by splenic and hepatic Fcγ baring macrophages. This view is supported by observations of decreased platelet survival when using Indium-111 tropolonate labelled autologous platelets transfused into ITP patients [61]. Since virtually all patients had reduced platelet survival, this would indicate that autoantibodies were present in all patients even if in many cases they remained undetectable. Platelet uptake by macrophages is unlikely to be the only mechanism by which autoantibodies trigger platelet consumption. There is increased complement fixation on platelets in the presence of autoantibodies [62]. Accumulation of a membrane attack complex contributes to phagocytosis but it can also lyse platelets directly.

The activity of autoantibodies could be determined by the target antigens. Autoantibodies against the GPIb/IX complex may have specific pathogenic actions and are less responsive to some therapeutic interventions [63, 64]. A multicentre study found significantly less response to intravenous immunoglobulin treatment in patients with anti GPIb/IX autoantibodies [65]. There is some evidence that an anti GPIbα autoantibody induced P-selectin and phosphatidylserine externalization on platelets, clustering of GPIbα on lipid rafts and activation of FcγRIIa [66]. It is known that localization of GPIb/IX to lipid rafts is required for interactions with FcγRIIa [67]. On the other hand, platelet activation determined as PAC-1 antibody binding to GPIIb/IIIa or by P-selectin expression was not detected in ITP platelets [68], therefore the role of platelet activation in ITP is yet to be fully established. The involvement of FcγRIIa in anti GPIIb/IIIa autoantibody activity is still to be clarified. Li et al. found that the activity of anti GPIIb/IIIa monoclonal antibodies was decreased by inhibition of FcγRIIa [69]. In these experiments, anti GPIbα antibodies were not affected by FcγRIIa neutralization [69]. Since these antibodies were generated in mice they may not emulate the behaviour of human ITP autoantibodies.

It has been proposed that there is a failing of peripheral tolerance that gives rise to ITP [47]. There is increasing evidence that dysregulation of T cells is part of ITP pathogenesis. Several studies have shown decreased number of CD4+ T cells in ITP [70, 71]. Impaired T cell activity can be temporarily restored by immune suppressive treatments such as dexamethasone [72] or with TPO mimetics [73]. Others found no differences in the frequency of T cells in ITP patients relative to controls [74, 75]. Similar frequency of T cells, however, may not reflect their activity and T cells from chronic ITP patients demonstrated reduced immunosupresive activity in vitro [74]. The functional role of Tregs was also shown in a murine model of ITP [76]. Overall, current evidence indicates that compromised Treg activity is part of a pathogenic progression that allows ITP to occur.

Recent Developments

Anti GPIb/IX antibodies have been shown to induce platelet desialylation (removal of sialic acid moieties from platelet glycoproteins, mainly GPIbα). Desialylated platelets are consumed by hepatocytes through interactions with the Ashwell-Morell receptor (AMR) (also known as asialoglycoprotein receptor or ASGPR). There are indications that desialylated platelets may also be targeted by cytotoxic T cells [77]. Platelet desialylation was detected in mice injected with anti mouse GPIbα antibodies and platelet clearance was reduced by inhibition of the AMR receptor or by prevention of desialylation using an enzymatic inhibitor [78]. A strong anti GPIb/IX autoantibody from an ITP patient was found to induce platelet activation and desialylation in vitro [79]. A study of 61 ITP patients found indications of desialylation, especially in those with anti GPIb/IX antibodies [80]. The differences, however, were minor and this has to be corroborated in additional studies, especially in view of reports that there is desialylation in platelets from patients with anti GPIIb/IIIa autoantibodies [77].

In an animal model of ITP using an anti GPIIb antibody, Leytin et al. showed that there was induction of platelet apoptosis as measured by mitochondrial membrane depolarization, PS exposure and caspase 3 activation [81]. Indeed, it had already been demonstrated that ITP platelets possessed apoptotic characteristics [82]. Soon after, seminal work by Mason et al. established that platelet life-span is controlled by apoptosis with critical involvement of pro-apoptotic factors Bak and Bax [83]. This implicates platelet apoptosis as a mechanism that may affect platelet survival in ITP. Recent work found markers of apoptosis in ITP platelets, in particular PS exposure and mitochondrial membrane depolarization [68]. The authors also showed that ITP serum or purified IgG induced mitochondrial membrane depolarization in healthy platelets. Caspase 3 activation could not be demonstrated in in vitro treated platelets [68], and this is consistent with other reports in which no caspase 3 activation was observed in treated MK [55]. Taken together, these studies indicate that platelet life-span reduction in ITP is in part due to increased platelet apoptosis.

The role of infection in the pathogenesis of ITP is becoming important and merits consideration. ITP patients are at higher risk of infection, but this may be due to the toxic effects of some therapies such as immunosuppressant drugs or corticosteroids or to decreased platelet numbers. The causative role of H. pylori in many cases of ITP has already been established. In a systematic review of 25 studies, Stasi et al. evaluated the platelet counts of 696 ITP patients after H. pylori eradication and found a complete response of 42.7% [84]. Eradication therapy appears to be more effective in patients with higher platelet counts. For those with <30 x 10⁹ platelets/L the overall response rate was 20%. In view of these high levels of response, these investigators have recommended testing of newly diagnosed ITP patients for H. pylori infection [84]. The cost effectiveness of H. pylori screening has to be considered. It is likely to be beneficial in regions with high prevalence of the infection such as Japan, but may be of little value in the United States where the prevalence is low. Nevertheless, H. pylori testing should be conducted in patients with refractory ITP since chronic infections can contribute to ITP pathology [85].

A Swedish study found that ITP patients had a higher incidence of infections (of bacterial, viral and fungal origin) relative to the general population within 5 years prior to ITP diagnosis [86]. This study excluded patients with secondary ITP or with ITP due directly to infections [86]. It is not known whether patients that eventually developed ITP had an already altered immune system that made them predisposed to infections or if infections contributed to the development of ITP. In some instances, activation of immune cells or molecular mimicry (such as in the case of the varicella zoster [87] or hepatitis C viruses [88]) may trigger development of ITP. In view of the fact that integrin-like beta-propeller domains are found in bacteria [89], molecular mimicry is likely to be a more important factor in the development of ITP than currently recognized. Even fungal infections with Candida albicans can progress into ITP [90], therefore the role of infection in ITP is an area that needs active exploration. A summary of the effect of autoantibodies on platelets is shown in Fig. (1).

Fig. (1))

Effect of autoantibodies on platelets and megakaryocytes. Failure to eliminate autoreactive clones is proposed to lead to autoantibody production. Interaction of autoantibodies with platelets leads to decreased platelet survival via complement fixation and cell lysis [62], phagocytosis [91], apoptosis [68] and desialylation [79, 80] while autoantibody/MK interactions can affect proliferation and differentiation [92], proplatelet formation [55, 93] and induce apoptotic changes [50, 94].

Impairment of Platelet Production in ITP

As described above, platelet production involves a series of steps that can be broadly described as MK development and platelet formation and release. Both aspects can be affected in ITP and as a consequence the number of functional platelets produced is compromised. MK express GPIb/IX and GPIIb/IIIa abundantly [92] and can be bound by anti platelet antibodies [95]. MK/antibody interactions have been proposed to lead to MK injury and interference with platelet production [96, 97]. Features consistent with MK apoptosis have been observed in bone marrow from ITP patients [50] and there is evidence that apoptotic changes are a feature that contributes to low platelet production in ITP [98].

Observations that platelet production was compromised in ITP patients [99] suggested that platelet destruction was not the only mechanism that accounted for low platelet counts in ITP. It was then shown that plasma from some ITP patients supressed MK production in vitro [100]. The relevance of inhibition of MK production in vivo is unclear since the majority of ITP patients appear to have normal numbers of MK in the bone marrow and only a minority may have increased, rather than reduced, MK numbers [101]. In addition, no difference in bone marrow assessment was observed between younger (<60 y.o.) and older ITP patients, thus supporting the concept that bone marrow examination is of little value for ITP diagnosis [102].

In this case, how might platelet production be affected in ITP patients? Iraqi et al. and Lev et al. demonstrated the deleterious effects of autoantibodies on proplatelet formation and platelet release in vitro using MK derived from cord blood [55, 93]. These authors found that ITP autoantibodies did not produce substantial apoptotic changes in MK and did not affect proliferation and differentiation significantly. Importantly, the effects on proplatelet formation [55, 93], proplatelet architecture [93] and platelet release in culture [55] are strong indications that autoantibodies effect principally the capacity of MK to produce platelets and not their proliferation and differentiation. The effect of an anti CD41 antibody on mouse MK function was also observed in a murine model of ITP [103].

The function of apoptosis in MK development and platelet production has a contestable history. A number of experiments indicated that a level of MK apoptosis was an intrinsic requirement for magakaryopoiesis and platelet production [104, 105] (see review by Li and Kuter [106] and by Perdomo et al. [98]). For instance, inhibition of caspase activity interferes with platelet formation [105] and overexpression of the anti-apoptotic factor Bcl-2 also caused a decrease in platelet numbers [107].

Recent work has indicated that MK apoptosis must in fact be controlled for proper function. Deletion of the crucial pro-apoptotic proteins Bak and Bax did not affect platelet production in vivo [108]. Conversely, mice lacking the pro-survival factor Bcl-xL had apoptotic MK that failed to produce platelets [108]. In addition, it has been shown that caspase 9 is the initiator caspase in the induction of MK apoptosis but that its absence does not affect platelet production or function [109]. In a recent study of ITP patients, it was found that MK from patients showed less signs of apoptosis than those from normal controls [109]. Therefore, genetic studies in mice indicate that induction of MK apoptosis in not required for platelet production and stimuli that alter apoptosis have substantial effects on platelet production.

Disease Characteristics and Therapies

ITP patients present with diverse symptoms. In some, signs of bleeding are common (mucosal bleeding, petechiae and ecchymoses), while others show no manifestations of bleeding even when the platelet counts are low. There is also a risk of fatal hemorrhage, especially intracranial bleeding. In a literature review of 1817 ITP patients, Cohen et al. reported 49 cases of fatal bleeding [110]. The most at risk patients were those with persistent low platelet counts (<30 x 10⁹/L). Bleeding risk also increases with age, the presence of other diseases and trauma.

Overall, ITP patients have a higher risk of death, predominantly from bleeding and infections [111] and also show a deficit in health-related quality of life [112]. Fatigue is an important factor in ITP [113], which combined with susceptibility to infection and medical visits results in work difficulties and reduced productivity [114].

The categorisation and definitions of ITP used here follow the recommendations of the international working group [44]. The classifications of ITP are described in Table 1.

Table 1 Definitions and terminology of ITP based on the international working group guidelines [44].

a A re-evaluation of the definition of refractory ITP has been proposed. This would include patients not responding to rituximab or to high doses of TPO receptor agonists or to rescue therapy, those that cannot undergo splenectomy and patients with affected quality of life [115].

Haemostasis is of primary concern in ITP, therefore the aim of therapy is to stabilize the platelet count and to reduce the risk of bleeding. The presence of bleeding and not the degree of thrombocytopenia determines the severity of ITP [116]. Reaching normal platelet counts should not be the objective. A platelet level of < 30 x 10⁹/L is generally considered the cut-off to commence treatment [117], however other considerations such as bleeding tendency and co-morbidities usually influence this decision. The reason to delay treatment unless