Haematology Research Unit, St George and Sutherland Clinical School, Faculty of Medicine, University of New South Wales, Sydney, NSW, Australia
Immune thrombocytopenia (ITP) is an autoimmune bleeding condition characterised by low platelet counts (< 100 x 109/L; reference count 150 – 400 x 109/L) in the absence of other triggers that may cause thrombocytopenia2. Bleeding (petechiae, mucosal bleeding) is a common sign of ITP. There is a risk of internal bleeding and fatal haemorrhage has been reported3. ITP patients have reduced quality of life and decreased life expectancy4. Autoantibodies against abundant platelets antigens such as glycoproteins (GP) Ib/IX and IIb/IIIa are found in ITP patients and contribute to platelet destruction. Platelet production by their precursor cells, megakaryocytes (MKs), is also decreased in ITP5. This is thought to occur via autoantibody-mediated detrimental effects on MK differentiation and proliferation, which subsequently affect platelet production. These conclusions were drawn from observations of MKs cultured in the presence of ITP plasma6, 7, antiplatelet monoclonal antibodies8 or serum from drug-induced ITP9. However, a recent investigation in a murine passive model of ITP showed that most anti-platelet antibodies did not produce changes in MK numbers or morphology10. It should be noted that this work10 used monoclonal or polyclonal antibodies against mouse platelets and not ITP patient’s-derived IgG. Nevertheless, as will be discussed below, the observations of this mouse model study are in agreement with our conclusions using ITP antibodies on human MKs in vitro1.
Treatment of ITP comprises several phases depending on severity and response. The goal of treatment is to decrease the risk of bleeding. The first line of therapy involves the use of corticosteroids or intravenous immunoglobulins. The objective of corticosteroid administration is to reduce antibody production11. In addition, steroid treatment also decreases platelet activation in responsive ITP patients12. Several mechanisms of action have been proposed for IVIg, including clearance of the autoantibodies mediated by FcRn13, 14 and induction of the Fcγ inhibitory receptor15. The evidence for these mechanisms has been challenged (reviewed by16), and involvement of the FcγRIIIa has been proposed16. These therapies can produce a rapid increase in platelet counts but do not produce a lasting response in most patients. Non-responsive patients can be treated with second line therapies such as splenectomy, which can be effective and long-lasting due to the removal of an organ that is involved in both platelet destruction and autoantibody production. Immunosuppressants such as anti CD20 (rituximab)17, cyclophosphamide18, cyclosporine or mycophenolate mofetil19, which inhibit B or T cells are also used as second line therapies. More recently thrombopoietin (TPO) receptor agonists (TPO-RAs) such as romiplostim and eltrombopag have entered clinical practice20. These agonists act by activating c-Mpl on precursor cells and promoting MK proliferation.
A well-recognised aspect of ITP is its heterogeneity, especially patients’ response to treatment. TPO-RAs, for instance have a durable response mainly in non-splenectomised patients21. In addition, long-lasting complete responses have been documented after discontinuation of TPO-RAs therapy22. Durable response is not a universal feature of these drugs and the nature of this immunomodulatory effect is yet to be described. The literature is still unclear about how these drugs increase platelet production. Leon et al showed an increase in MK proliferation (which is expected for TPO-RAs) leading to enhanced platelet production23, while other authors documented increased platelet numbers despite no changes in MK mass24.
Our study1 used MKs derived from cord blood CD34+ cells to address two aspects of ITP, i) the effect on ITP patients’ serum and purified IgG on MK proliferation, maturation, apoptosis, proplatelet formation and platelet release; and ii) the activity of TPO-RAs on proplatelet formation in the presence of ITP autoantibodies. The findings of the study1 are summarised in Table 1. The principal observation was that most ITP autoantibodies inhibited proplatelet formation and this was unrelated to effects on MK proliferation, differentiation, apoptosis or ploidy status. This implies that it is the final stage of megakaryopoiesis that is targeted by ITP antibodies, most likely by the effects of these antibodies on proplatelet structure25. Three serum samples affected MKs in culture but this was unrelated to the presence of anti-platelet antibodies and was most likely due to other components present in these sera such as TGF-?? , PF4 or s-FAS1. In vivo, other mechanisms such as T cell imbalance are involved in ITP and may lead to reduction of MK production (reviewed by26), however in this case the experiments were performed in vitro in the absence of other cells.
Table 1: Effect of ITP serum or IgG on cultured MKs
|19 ITP Patientsa|
|Effectb||Serumc||P valuee||IgGd||P valuee|
|Decrease in Proplatelet formation||8 (42%)||<0.0001||13 (68%)||<0.0001|
|Reduction of MK proliferation||3 (15.7%)||0.0005||0|
|Reduction of MK maturation||3 (15.7%)||<0.0001||0|
|Decrease in cell size||3 (15.7%)||<0.0001||0|
Historically, it was thought that ITP was characterized by platelet destruction, but it is now apparent that insufficient platelet production is a substantial feature of ITP pathogenesis. Our work shows that IgG from most patients decreases proplatelet formation and consequently platelet production without evident effects on other MK features such ploidy, proliferation, cell size and apoptosis1. Damage to proplatelet structures by ITP antibodies in the absence of obvious apoptotic changes was also documented in a study by Lev et al25, which was conducted mostly concurrently with our work and is also supported by observations of normal MK numbers and morphology in a mouse model of ITP10. More recently, it was shown that thrombocytopenic patients (ITP and myelodysplastic syndrome) had reduced rather than increased MK apoptosis, which corroborates the in vitro observation of no significant apoptosis induction upon ITP antibody treatment.
TPO-RAs represent a relatively effective therapy for patients refractory to first line treatments. This in turn increases platelet production. Unexpectedly, MK proliferation has been observed in both responders and non-responders to TPO-RA administration27, suggesting that in some patients there is no direct correlation between MK proliferation and increase in platelet counts. Our work1 offers an explanation by the surprizing finding that TPO-RAs not only promote MK proliferation but can also neutralize the inhibitory activity of some ITP antibodies, thus increasing the number of proplatelet-bearing MKs. In addition, the response to the TPO-RAs used in our study1 (romiplostim and eltrombopag) varied among patient samples, suggesting that some non-responders might benefit by switching TPO-RAs. This might become more pertinent once additional TPO-RAs such as avatrombopag28 become widely available. This study does not question the view that platelet destruction is also a key feature of ITP, as documented many decades ago29. In fact, several mechanisms, including phagocytosis of opsonised platelets by macrophages30, complement-mediated destruction31, apoptosis and desialylation32 may account for platelet depletion in ITP. Liu et al reported that TPO-RAs may also inhibit platelet destruction by upregulation of inhibitory FcγRII and suppression of activating Fcγ receptors33. This expands the potential range of TPO-RA activities to MK proliferation, increase of pro-platelet formation per MK and inhibition of platelet destruction. In conclusion, it is evident that in both platelet destruction and production are contributors to the development of ITP. Our work shows that megakaryopoiesis is largely unaffected, but antibody-mediated inhibition of proplatelet formation and platelet release is a regular feature of the pathogenesis of ITP.