Answers are grounded in this paper and cite their sources.
Autoimmune disorders, in which there is loss of tolerance to self antigens, affects 5% of the population representing 14 million individuals in the USA and 3 million in the UK (Editorial, 2001). To date, the treatment of these disorders in adults has been disappointing, in part because of a lack of understanding of the underlying pathophysiology in this diverse group of disorders. Some 50 years ago, Harrington et al (1951) elegantly demonstrated the presence of a 'factor' in the globulin fraction in the peripheral blood of patients with idiopathic thrombocytopenic purpura (ITP) that caused the thrombocytopenia. We now know that this factor was anti-platelet antibody. However, despite major advances in our understanding of immunology and autoimmunity, these have not been translated into improved diagnostic or therapeutic strategies for diseases such as ITP. Although several different classes of drug and other forms of treatment have been used in chronic ITP, no therapeutic modality is truly evidence based (George et al, 1996). In our opinion, ITP represents an ideal model of organ-specific autoimmune disease for study. It is sufficiently common to make its study feasible and has the advantage of being manifested by a reduced peripheral blood platelet count. As such, the efficacy of treatment can be objectively measured by using the full blood count. In this article, we look at how far we have moved on since Harrington's seminal work and define the disorder, discuss the underlying immunological features, summarize the standard therapy and evaluate the treatments available for non-pregnant adults who are refractory to standard first-line therapy. We have also reviewed therapies under development that appear to offer a more targeted approach to treatment. Using ITP as a model of human autoimmune disease could logically lead to improvements in the management of more complex disorders such as multiple sclerosis, systemic lupus erythematosus (SLE) and type I diabetes mellitus. Idiopathic thrombocytopenic purpura (ITP) is an autoimmune disorder in which platelets, opsonized with anti-platelet autoantibodies, are removed prematurely by the reticuloendothelial system (RES), leading to a reduced peripheral blood platelet count. Although bone marrow megakaryocytes are often increased, relative marrow failure may play a role in a proportion of patients (Ballem et al, 1987). The aetiology of ITP in adults is unknown and the clinical course is variable and unpredictable. The incidence of chronic adult ITP is around 5·8–6·6 new cases per 100,000 population per year in the USA (Bottiger & Westerholm, 1972; McMillan, 1997), and a similar incidence in the UK. The overall combined incidence of new cases of adult and childhood ITP is higher, and may reach 20 per 100,000 population per year. The pathogenesis (and treatment) of ITP has been reviewed by several authors (Berchtold & McMillan, 1989; Chong, 1995; Ikehara et al, 1995; Semple & Freedman, 1995; Karpatkin, 1997; Wang & Shen, 1997; Blanchette et al, 1998; Yang & Zhong, 2000; Lechner, 2001). ITP in childhood is generally termed 'acute' as the illness is seasonal, typically follows a trivial viral infection or vaccination, and in most cases is transient, requires notreatment and spontaneously recovers. In the adult (chronic) form there is usually no obvious antecedent illness and most patients have chronic thrombocytopenia; spontaneous recovery is uncommon (George et al, 1996). In adults, the frequency of death from haemorrhage in patients failing to achieve an adequate platelet count is 5% (George et al, 1996), although the frequency may be influenced by age and duration of thrombocytopenia (Frederiksen & Schmidt, 1999). The ITP 'phenotype' is heterogeneous: some patients suffer major bleeding from the outset, while others have few problems apart from an increased bruising tendency. This may partly be explained by the acquired platelet dysfunction which is seen in some patients with ITP, which in turn may be related to the target antigen involved in the autoimmune process (this is discussed later in Autoantibody characteristics). Autoantibodies reacting with glycoprotein (GP) IIb/IIIa affect platelet aggregation and anti-GPIb/IXautoantibodies impair platelet adhesion to the subendothelial matrix, causing unexpectedly severe bleeding for the level of the platelet count (Wang & Shen, 1997). In general, however, in contrast with thrombocytopenia due to marrow infiltration (e.g. leukaemias, lymphomas or other malignancies) or aplasia, patients with ITP are able to tolerate remarkably low platelet counts and maintain an adequate quality of life (Karpatkin, 1985). The degree of bleeding is largely dependent on the platelet count and patients with platelet counts below 10 × 109/l (and usually below 5 × 109/l) are at greatest risk of bleeding, including intracranial bleeding. Basic investigations comprise: clinical history and physical examination, full blood count, blood film examination, and bone marrow aspiration; the latter is controversial, and in childhood ITP marrow examination is not carried out unless there are atypical features (Eden & Lilleyman, 1992; Lilleyman, 1999; Medeiros & Buchanan, 2000). In adults with ITP, the American Society of Hematology panel consensus view was that marrow examination should be performed in patients over the age of 60 years. Myelodysplasia, which may present with thrombocytopenia in this age group, requires exclusion and bone marrow examination is appropriate. Other tests such as the autoimmune screen aim to exclude other autoimmune diseases which may be associated with thrombocytopenia, e.g., SLE or lupus anticoagulant/antiphospholipid syndrome. In a small proportion of cases, there may be a familial cause and this should be elicited from the history. Despite the availability of a variety of immunological assays (e.g. anti-platelet antibodies, direct platelet immunofluorescence test, monoclonal antibody immobilization of platelet antigens and others) the diagnosis remains clinical and one of exclusion. The generation of autoreactive T cells has been shown to be a feature of the normal immune system and is likely to be beneficial, and required for normal tissue repair (Schwartz & Cohen, 2000). However, a number of studies have shown strong evidence for dysregulation of the immune response in autoimmune disease (Refaeli et al, 1999; Straus et al, 1999; Wucherpfennig & Eisenbarth, 2001); there may be failure of the normal checkpoints such as clonal deletion, anergy, immune deviation and suppression which control autoreactive T cells in health (Stockinger, 1999). From studies of autoimmune disease, it is clear that ITP and other autoimmune disorders are multifactorial. It is likely that loss of tolerance to a self antigen alone is insufficient togenerate the autoimmune disorder (Mackay, 2000); instead, patients probably require: (i) a specific set of genetic determinants [e.g. polymorphisms within major histocompatibility complex (MHC), cytotoxic T-lymphocyte antigen 4 (CTLA4) or other genes]; (ii) dysregulation of the immune response (involving dendritic cells, T or B cells, or all three); (iii) an environmental 'trigger'. Autoimmune disease arises only when all these determinants are present in an individual at one particular time. This is further reinforced by the observation that self-reactive lymphocytes are commonly found in normal individuals. For example, siblings of patients who have autoimmune disorders are more likely to have autoantibodies themselves, albeit at lower titres than their affected siblings, without evidence of autoimmune disease per se, perhaps because they have not been exposed to the environmental trigger required to tip the balance toward autoimmune disease. Genes clearly play a part in the development of autoimmune disease, but other interactions are of major importance. Evidence for this comes from the observations of identical twins, one of whom has autoimmune disease while the other does not, i.e. there is a lack of concordance. In pairs of monozygotic twins where one twin has autoimmune disease, the other twin is affected in less than 50% of cases (Ermann & Fathman, 2001). Genetic studies over the years have concentrated on determining the contribution of MHC genes to the development of autoimmune disease, and strong associations have been reported between particular MHC haplotypes and disease. The best example is ankylosing spondylitis, associated with human leucocyte antigen (HLA) B27, which has a relative risk of 87·4. HLA studies in ITP have shown an increased frequency of HLA-DRB1*0410 in Japanese adults with ITP (Nomura et al, 1998). Conversely, the possession of a disease haplotype does not predict disease and studies of rheumatoid arthritis (RA) have shown that RA can develop in patients lacking a typical disease-associated MHC haplotype (Feldmann, 2001). In other diseases such as insulin-dependent diabetes mellitus (IDDM), the MHC association is so strong that the MHC alleles can be used as markers of predisposition to IDDM. How do allelic variants within MHC genes predispose to diseases such as IDDM? Autoimmune responses involve T cells and it is not surprising that there are strong MHC associations (Gautam et al, 1994; Haines et al, 1996). MHC polymorphisms may be associated with variable ability of MHC molecules to present autoantigenic peptides to T cells (Ridgway & Fathman, 1998). From studies of humans with IDDM and non-obese diabetic (NOD) mice with spontaneous IDDM, it has been shown that there is substitution of the negatively charged amino acid aspartic acid with a neutrally charged amino acid (Morel et al, 1999) which may influence antigen presentation. Rheumatoid HLA-DR variants that are associated with an increased risk of disease or are associated with more severe phenotypes contain a shared epitope, in the form of a conserved amino acid sequence within the β chain of the MHC class II molecule. Acute ITP in children is often preceded, and presumably triggered, by a viral illness or immunization, but which environmental triggers might be involved in adult ITP? Studies of patients with multiple sclerosis (MS) have shown that there are features of infectious disease (Steinman, 2001). In addition, previous studies involving patients with SLE have shown a high incidence of prior infection with Epstein–Barr virus (EBV). A single antigenic determinant on EBV has been shown to be shared with one particular SLE autoantigen (James & Harley, 1998). Many patients with ITP have elevated levels of platelet-associated immunoglobulin (Ig)G which may represent the anti-platelet antibody. However, platelet-associated IgG rises in other non-immunological causes of thrombocytopenia (George & Saucerman, 1988). The autoantibodies involved in ITP are generally IgG, but IgA and IgM autoantibodies have also been reported (Kiefel et al, 1996). Using antigen-specific assays such as the monoclonal antibody-specific immobilization of platelet antigens (MAIPA), platelet-associated IgG and antigen capture assays, several platelet target antigens have been characterized. These include GPIIb/IIIa (αIIIbβ3, the fibrinogen receptor) and GPIb/IX (the von Willebrand receptor) which appear to be the most frequently involved (Brighton et al, 1996; Macchi et al, 1996; Warner & Kelton, 1997; Warner et al, 1999). Less commonly GPIa/IIa, GPIV and GPV are involved. Recent reports suggest possibly 40% of autoantibodies are reactive to both GPIIb/IIIa and GPIb/IX (Stockelberg et al, 1995), possibly due to the serum in some patients with ITP containing two different IgG antibodies. In terms of disease chronicity, GP-specific autoantibodies may be important in the pathogenesis of chronic ITP (Hou et al, 1995); from available data, GPIIb/IIIa appear to play a major role in the development of chronic ITP in 30–40% of cases (McMillan et al, 1987; Kiefel et al, 1991). Opsonized platelets are removed prematurely by the reticuloendothelial system (RES) predominantly through an Fc-dependent mechanism (Garvey, 1998); Fc-independent effectors may also play a role. Recently, Nieswandt et al (2000) have looked at the pathogenic effects of IgG monoclonal antibodies (mAbs) of different IgG subclasses against murine GPIIb/IIIa, Iba, Ib/IX, V and CD31. Their data suggest that, at least in mice, the antigenic specificity of the anti-platelet antibodies determines the pathogenic effects rather than the IgG subclass. They also demonstrated that antibodies against GPIb/IX caused thrombocytopenia through an Fc-independent mechanism, while that from autoantibodies against GPIIb/IIIa involved the Fc system. Further work is clearly needed in order to determine the significance of all of these findings, which may translate into stratification of patients into those in whom Fc receptor blockade or inactivation is a useful option, and those in whom it is not. Before discussing treatment options, it is useful to examine the long-term outcome of adults with ITP. Despite many studies of various treatment options, there have been few reviews of the outcome of cohorts of adult patients with ITP to determine whether ITP results in an increase in morbidity or mortality compared with that of the general population. A recent natural history study by Portielje et al (2001) looked at the outcomes of 152 adults with ITP. They were able to show that 93% of patients ultimately achieved platelet counts greater than 30 × 109/l and did so within 2 years of diagnosis. They also showed that 85% of patients achieved platelet counts above 30 × 109/l off treatment andhad an overall long-term mortality similar to that ofthegeneral population. Some 9% of patients with severethrombocytopenia (platelet counts < 30 × 109/l) had refractory disease and had a mortality risk fourfold higher than the general population. Deaths in this group were due equally to infection and bleeding. Six per cent of their patients required maintenance therapy to achieve a sustained platelet count above 30 × 109/l. From this single study, it would appear that in most adults with moderate ITP the disorder runs a benign course, while the severe refractory group have higher morbidity and mortality than the general population. As very few trials have been conducted in ITP, we do not know which constitutes the best treatment. In addition, the efficacy of the available therapies is unpredictable and often transient. With some treatments, the side-effects may be severe and this must be balanced with the likely benefits. In general, patients with platelet counts exceeding 30 × 109/l require no treatment unless they are undergoing any procedure likely to induce blood loss, including surgery, dental extraction or delivery (Yang & Zhong, 2000). Standard therapy comprises oral corticosteroids, intravenous immunoglobulin (IVIg), and splenectomy. Around two thirds of patients will respond to prednisolone at the traditional dose of 1 mg/kg body weight/d tapering off over several weeks, but relapse is common when the dose is reduced. Both lower (0·5 mg/kg/d) and higher (2 mg/kg/d) doses appear equally effective (Mazzucconi et al, 1985; Blanchette et al, 1998). Around 20–30% can expect a long-term response (Berchtold & McMillan, 1989; Manoharan, 1991; George et al, 1996; Blanchette et al, 1998). Pooled normal human immunoglobulin is effective in elevating the platelet count in 75% of patients, of which 50% will achieve normal platelet counts. However, 3–4 weeks following IVIg treatment platelet counts drift back to pretreatment levels (George et al, 1996). The effects of IVIg are usually not sustained, although some long-term responses have been reported (Newland et al, 1983). The mechanism of action of IVIg in ITP remains largely unknown but involves Fc receptor blockade, anti-idiotype antibodies in IVIg which block autoantibody binding to circulating platelets and immune suppression (Godeau et al, 1993; Chong, 1995). IVIg is well tolerated with few major side-effects. Splenectomy is the only curative treatment modality in ITP and was used before steroid therapy was introduced in 1950 (Chong, 1995; George et al, 1996). Two thirds of patients with ITP who undergo splenectomy will achieve a normal platelet count which is often sustained with no additional therapy. Patients who do not have a complete response can still expect some improvement in counts (e.g. partial response) or transient increases in platelet count (George et al, 1996). Recent work using indium-labelled autologous platelets has shown that when platelet destruction is splenic then 96% of patients between the age of 5–30 years and 91% of those above the age of 30 years can expect to obtain a remission. However, where the platelet destruction was hepatic or diffuse (mixed splenic and hepatic), 92% of patients failed to normalize their platelet counts or had incomplete responses to splenectomy (Najean et al, 1997; Cavenagh et al, 1999). Indium scanning therefore appears to be a reliable predictor of response to splenectomy, although others have proposed the response to IVIg as a predictor of likelihood of success of splenectomy (Law et al, 1997). This defines those patients who fail to respond to standard treatment or require unacceptably high doses of corticosteroids to maintain a safe platelet count. A number of drugs have been used as second-line therapy for ITP and those used will depend on the age of the patient, the severity of the presentation, the platelet count, whether the disease is primary refractory or relapsed, and the length of time prior to relapse. When considering second-line therapy, the first-line therapies (corticosteroids, reticuloendothelial blockade with IVIg or splenectomy) should be considered and implemented if possible (George et al, 1996). The doses may have to be altered compared with those used in first-line therapy and additional supportive therapy may need to be considered to allow their chronic use, e.g. there is a small but significant risk of osteoporosis and avascular in patients on long-term corticosteroids, and they should be for this and with or if The for treatment should be the and side-effects of against the of no treatment. In some patients, therapy such as may be if there is only bleeding, and in severe bleeding, such as platelets may be Although these are from the and have a they are often effective in bleeding et al, For the in whom further treatment with corticosteroids is there variety of therapeutic & These (i) (ii) (iii) intravenous including and and The variety of treatments available for second-line therapy their relative lack of efficacy and treatment should be to the It also in that for the who has following splenectomy, an may be present in to and this should be looked for using while considering further treatment. As an to reported responses in refractory patients using an oral of for 4 for patients were in this small study with responses in all patients had platelet counts exceeding × 109/l) sustained for at least this side-effects are common and studies conducted by other have not with such success et al, 1995; & 1996; et al, 1997). success has also been reported using at a lower dose such as 20 for 4 such as have been used as and treatments for patients with refractory ITP. study reported the results of adult patients with platelets less than 50 × all of whom were with oral corticosteroids at 1 was at 30 for 3 20 for 4 then 2 and 1 for 1 The platelet count normal within in all patients, in of the response only a few weeks before to pretreatment levels et al, 1991). von et al compared the of with IVIg in adult patients with a control patients with standard oral The was found to be as effective as IVIg in terms of the frequency of with no reported side-effects. The major between the was that the response to oral was than to intravenous the response to intravenous was transient in all patients and maintenance with oral was required to maintain an adequate platelet count. intravenous immunoglobulin at a dose of 1 for 2 often in with corticosteroids, will the platelet count in a proportion of patients but may be associated with significant in particular but if can be on an or with intravenous et al, has been shown to the platelet count in of adults et al, 1997). The mechanism of action is to be through destruction of cells which are removed by the reticuloendothelial In a single study of in and patients, of patients showed an increase in platelet count of greater than 20 × 109/l and in of patients the platelet count by more than 50 × 109/l. The improvement for more than 3 weeks in 50% of patients who et al, 1997). treatment is for patients who are not but is not for refractory patients following splenectomy. This group of drugs may cause a transient increase in the platelet count between 1 and 3 weeks in of patients responses have been in less than of patients (Berchtold & McMillan, 1989; Manoharan, 1991; George et al, 1996; Blanchette et al, 1998). this treatment is although it may be of when other therapies have failed and there is a need to the platelet count an appears to be effective in patients with between ITP and and can often be used as a in patients but who require unacceptably high et al reported the outcome of patients, of which had splenectomy, with at a dose of for more than 2 Around showed of the platelet count above 50 × 109/l sustained from more than 2 patients and those who have splenectomy to respond The mechanism of is unknown but is to the number of Fc on splenic platelet et al, 1997). may be required in patients who fail to respond to with or as single may be considered and to of patients may have a sustained However, because of including which are more common in patients with these of cytotoxic therapy, in those who are should be In those patients with chronic ITP refractory to multiple previous treatments, the of and prednisolone combined in such as those used in the lymphomas have been shown to be et al reported their results in 10 patients with refractory ITP. Two patients had underlying disease disease and chronic 10 had been with and had splenectomy. The platelet count was less than 5 × 109/l in all A complete response was seen in patients in a partial response was in two and two of In a of adults with chronic ITP with at responses were in with a duration of treatment required to achieve a response of responses were in patients (Godeau et al, 1997). mechanism of action is unknown but may be due to reticuloendothelial blockade through increased destruction (Godeau et al, 1997; et al, 1999). to of adults with ITP fail to respond to or second-line therapies and major Some of these are (e.g. but many are only with bleeding, a management there have been few trials the efficacy of and second-line treatments, the when one at the evidence for In these represent a group of often with in common apart from the that in some patients these drugs may the platelet count. are no quality trials for any of these therapies and there are no useful of it for the to know which to most adult patients with chronic refractory ITP are able to tolerate thrombocytopenia well et al, and are able to have normal or normal quality of For those who fail to respond to standard and second-line therapy and who require the are and (i) no drug (ii) (iii) A and other treatments such as or two are discussed This of treatment is more largely because we often all therapeutic to no and are not out in patients who are largely Many of the in are associated with significant when used in high From studies of childhood ITP, we know that despite patients have no major bleeding they can tolerate very low platelet counts and still maintain a normal quality of will have patients with severe refractory ITP in whom all treatments have been and who are on no treatment and This management also has some it is and and the patients are a full of they are not to be any further therapy, they are usually We need to that with our of there will be some patients for whom there is no effective therapy. patients should be under and to any or to the or A has been shown to increase the platelet count when alone or with appears to be levels and are studies involving the drug have been conducted but from available data it may the over a are several that on the of in refractory ITP and have shown that of patients can achieve a platelet count of over × 109/l for between 1 and the treatment et al, 1989; George et al, 1996). However, is probably of in severe thrombocytopenia, and in some cases may the thrombocytopenia and induce bleeding et al, The mechanism of action is unknown but may be due to of B lymphocytes involved in the autoimmune of and of most and patients are to et al, 1998). et al reported on the of A in patients with refractory ITP, of whom had splenectomy. patients were treatments for of the were on steroid < 30 Some of patients had responses exceeding × while had responses between 50 and × the patients had The mechanism of action of the A may involve platelet et al, 1998). The of A requires
No takes yet. Share an insight, caveat, or question.
Provan et al. (2002) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: