Alternate considerations for current concepts in ITP

Abstract

Immune thrombocytopenia (ITP) is one of the most common forms of autoimmune disease affecting both adults and children. In recent years, there have been tremendous developments in the understanding of the pathogenesis and treatment of this condition. However, certain concepts related to ITP are worth consideration in view of alternative explanations and evidence available. These include (i) ITP is a disorder where thrombocytopenia is induced by autoantibodies against platelets or megakaryocytes, (ii) the mechanism of action of corticosteroids in ITP is through suppression of these autoantibodies, (iii) splenectomy is effective in ITP since spleen is the site of platelet destruction, and (iv) splenectomized ITP patients are at a major risk of infections.

Introduction

The last decade has been a tremendous boon for patients with immune thrombocytopenia (ITP) with significant developments in the understanding of immune pathogenesis of this condition, and the advent of thrombopoeitic agents, which has revolutionized the treatment of refractory ITP. Despite this, the diagnosis of ITP remains one of exclusion and corticosteroids are still one of the most effective first-line treatments available for ITP. In addition, although surgical removal of spleen is less often chosen by patients and physicians as a treatment option nowadays, it remains the only cure for ITP and is practised widely in the resource-limited countries and on rare occasions in the developed nations. In this article, certain concepts which are embedded in the literature about the pathogenesis of ITP and its treatment with steroids and splenectomy are challenged.

ITP is caused by the immune destruction of platelets (and megakaryocytes)

Current concept

Several articles by international experts in ITP have commented on the immune pathogenesis of ITP. Until recently, ITP was considered a bleeding disorder in which autoantibodies are directed against the individual's own platelets, resulting in increased platelet destruction by the reticuloendothelial system.^1,2 In the last decade, different investigators came up with enough evidence for the important role of T cells in ITP.^3,4 Cytotoxic autoreactive T cells have been shown to play a part in at least some patients as demonstrated by Olsson et al.⁵ More recently, the role of megakaryocytes and the contribution of decreased platelet production were explained by McMillan and colleagues.⁶ In all these cases, the basic mechanism described is the elaboration of antibodies by the immune system and ensuing destruction of megakaryocytes or platelets.

Alternate concept

Despite the above studies, one of the overlooked pathophysiological features contributing to thrombocytopenia of ITP is the phenomenon of platelet aggregation. Increased platelet aggregation is a well-established cause of thrombocytopenia in several conditions like microangiopathic haemolytic anaemia, and heparin-induced thrombocytopenia. In the classical microangiopathic haemolytic disease, thrombotic thrombocytopenic purpura, deficient activity of the important enzyme, ADAMTS-13 (or the Von Willebrand factor cleaving protease) has been proven to be the cause of platelet aggregation.⁷ With regard to ITP, Moore et al.⁸ measured the ADAMTS-13 enzyme levels in a number of patients with TTP, and ITP, among others. While reduced protease activity was observed in 9 out of 20 patients with TTP, a reduced activity was also observed in 6 out of 20 patients with ITP. Another line of evidence for the role of ADAMTS-13 comes from the fact that thrombocytopenia in ITP is well known to be exacerbated by infections and inflammatory states. In this regard, ADAMTS13 activity has been shown to be drastically reduced in patients with inflammatory diseases characterized by CRP levels higher than 50 mg/l.^9,10 ITP is occasionally noted to be secondary to autoimmune conditions like systemic lupus erythematosis (SLE). Mannucci et al.¹¹ demonstrated lower levels of ADAMTS-13 in patients with SLE, although this paper did not examine the correlation with thrombocytopenia with ADAMTS-13. In this context, it needs to be remembered that, even if ADAMTS 13 activity is decreased in a substantial proportion of patients with thrombocytopenia of various causes including ITP, none have less than 10% as observed in thrombotic thrombocytopenic purpura patients.¹² As such, the significance of a moderately decreased ADAMTS 13 activity is still unknown.

Thrombosis is an increasingly recognized complication in ITP, although the exact mechanism by which the risk of thrombosis is increased in ITP is not yet clear.¹³ Both arterial and venous thrombotic disorders have been described in ITP although arterial clots are more commonly observed where platelet aggregation is thought to be the predominant pathophysiological factor.¹⁴ Several hypotheses have been put forward including the activity of antiphospholipid antibodies, steroid-induced increased thrombin generation, complement activation, nitric oxide depletion, and ADAMTS-13 deficiency.^14,15 A recent abstract at the American Society of Haematology meeting examined the relationship between Von Willebrand factor antigen and ADAMTS-13 as a contributor to the thrombotic risk in ITP.¹⁶ The authors concluded that an increased ratio of Von Willebrand factor antigen to ADAMTS13 may be contributory in the pathogenesis of thrombosis, in particular in patients with acute and unstable ITP.

One of the interesting aspects of the alternate concept is the possible explanation for the variable bleeding patterns in patients with ITP. It is well-known that some patients with ITP do not exhibit much bleeding, despite low platelets counts, the level at which some other patients tend to have moderate to severe bleeding. In most cases, bleeding is unusual if the platelet counts are more than 30 × 10⁹/L. Several explanations have been put forward to explain this observation mainly based on the presence of specific anti-platelet antibodies.

Firstly, there is a study of 409 patients who were tested for the presence of platelet-associated autoantibodies by direct-platelet immunofluorescence test and plasma antibodies directed against GPIIb/IIIa, GPIb, and GPIa/IIa by monoclonal antibody immobilization of platelet antigens. Those patients with platelet autoantibodies were observed to have the more chronic form of disease (83.5 vs. 68.5%) and severe symptoms of haemorrhagic diathesis (17.3 vs. 6.9%).¹⁷ Nomura et al. showed among 101 ITP patients, those who had anti-GPIb autoantibodies (19 patients), thrombocytopenia was more severe.¹⁸ These patients also had severe purpura and poor response to prednisolone. This has prompted Fabris et al.¹⁹ to use platelet autoantibodies as a prognostic marker in ITP more recently. The abnormal platelet function or thrombopathy in patients with platelet auto-antibodies has been also suggested to contribute to the bleeding diathesis, irrespective of the platelet count. It has been speculated that GpIIb/IIIa or GpIb/IX antibodies inhibit their respective binding to fibrinogen or Von Willebrand factor and thus alter the process of platelet aggregation.^20,21 This has been demonstrated in the description of pseudo-Glanzmann syndrome where patients have haemorrhage seizure without thrombocytopenia.²² Laboratory evidence for platelet aggregation problems have been described in four out of eight patients with ITP by Heyns et al..²³ In addition to the important role of autoantibodies as predictive of bleeding risk, it may also be hypothesized that the non-bleeders have higher level of active Von Willebrand factor activity making their platelets aggregate better, while the ‘bleeders’ have lesser activity. An interesting prospective study in this regard would be to correlate the bleeding scores in ITP patients with their blood groups since it is well recognized that O blood group individuals tend to have lower Von Willebrand factor activity compared to those with the other blood groups.²⁴

In summary, ITP is a condition where thrombocytopenia is due to immune destruction of platelets or megakaryocytes although platelet aggregation may contribute to the lower platelet count and the unusual occurrence of thrombosis and varying bleeding patterns in this condition.

Prednisolone is helpful in ITP by decreasing platelet destruction

Current concept

First-line ITP therapy includes corticosteroids, which are thought to globally influence the immune system by functionally suppressing antibody production, influencing T- and B-cell reactivity and affecting cytokines.²⁵

Alternate concept

Corticosteroids raise the platelet count in ITP also by influencing platelet production. Probably the best evidence for this concept comes from the studies from Gernsheimer et al.²⁶ who determined the mechanisms of platelet count increase after ITP therapy. Survival time and localization of radiolabeled autologous platelets and platelet-associated immunoglobulin levels were measured before and after prednisone therapy or splenectomy. Eleven of 12 patients (92%) responded to prednisone and were noted to have increased platelet production with no difference in platelet survival. Interestingly, in the patient who failed steroids, the pre-treatment platelet production was already supra-normal. Houwerzijil et al.²⁷ examined the predictive value of clinical and platelet kinetic parameters for treatment outcome in 75 patients with platelets ≤20 × 10⁹/l. The platelet kinetic studies showed that the platelet production rate was decreased, normal, or increased in 33, 48, and 19% of patients, respectively. Of all the patients who received prednisone at 1 mg/kg/day dosage, durable complete and partial response rate was noticed in 64% of the patients with decreased platelet production rate, compared to 34% of the patients with normal or increased rate (P = 0.03). It may be argued here that immunomodulatory properties of corticosteroids may still be the key mechanistic factor here, whereby the increased platelet production occurs secondary to the suppression of antibodies which may inhibit platelet production. In this context, a very recent study looked at the mechanisms of platelet increase by intravenous immunoglobulin and thrombopoieitin mimetics.²⁸ The investigators used immature platelet fraction in assessing treatment effects in ITP patients and provided evidence to support inhibition of platelet destruction as the primary mechanism of intravenous anti-D and IVIG, although intravenous immunoglobulins (IVIG) was also noted to enhance thrombopoiesis. Non-responders to thrombopoietic agents had increased megakaryocytes but not increased immature platelet fraction, suggesting that antibodies blocked platelet release from their precursors. Azathioprine as an additional immunosuppressive agent to inhibit autoantibody production was suggested in the accompanying editorial as a strategy to overcome this resistance.²⁹ Unpublished reports have shown some experience in this regard in the use of prednisolone as an adjunct to thrompoieitic agents in cases with poor response to the latter.

Another ‘useful’ effect of prednisolone in ITP patients is its effects on capillary endothelial integrity. It is well-known to the ITP physicians that an improved response in relation to bleeding is noticeable in those commenced on steroids much before the platelet count rises in ITP patients. It has been demonstrated in several laboratories that platelets play a crucial role in maintaining the endothelial integrity.³⁰ A resulting clinical implication of this ‘protective’ role of platelets is the capillary leakage characteristics of sepsis-induced thrombocytopenia.³¹ As early as 1980s, thinning of capillary endothelium in experimental and clinical thrombocytopenia was reported.^32,33 The reduction in mean thickness of the capillary endothelium supplying either skin or skeletal muscle reverted to normal after four days of prednisone administration even though at this time point the degree of thrombocytopenia was still severe. The possible explanations for this steroid effect include decreased vessel wall prostacylin production in vivo and significant alterations in protein synthesis by endothelial cells.^34,35

Alternate explanation given for the variable bleeding patterns in ITP patients is the presence or absence of associated inflammation. In other words, in an otherwise clinically well person (without associated inflammation or infection), the likelihood of bleeding is not high and vice versa. This lack of correlation between platelet counts and bleeding has been also demonstrated in murine models where the binding of anti-platelet immunoglobulin to endothelial cells of small vessels and subsequent antibody-mediated endothelial activation, was noted to be relevant for the causation of bleeding.³⁶ This was confirmed more recently by the group from Wagner labs who showed that inflammation may cause life-threatening haemorrhage during thrombocytopenia.³⁷ Both these studies demonstrate that platelets are required for the maintenance of vascular integrity in inflammation, and thrombocytopenia alone is not sufficient to cause spontaneous haemorrhage. Since prednisolone is a potent anti-inflammatory drug, the mechanism of action of these agents in ITP will also include the immunosuppressive functions. It is not yet demonstrated that these steroid effects are specific to prednisolone or may also be exhibited by dexamethasone.

In summary, prednisolone is helpful in ITP by decreasing platelet antibody production but also manages to reduce the bleeding episodes by maintaining the vascular integrity and acting as a potent anti-inflammatory agent.

Splenectomy is effective in ITP by removing the organ site of platelet destruction

Current concept

Since Kaznelon first demonstrated the therapeutic value of splenectomv in ITP patients in the early part of 20th century, an important role has been attributed to spleen in the pathogenesis of this disease. However, the exact mechanism by which spleen partakes in ITP pathophysiology is not yet entirely clear. Direct lysis, immune-mediated phagocytosis, and the production of antibodies are the accepted explanations.²⁶ For this reason, indium-labelled studies to analyse the site of platelet destruction have been recommended in some countries before considering splenectomy with the idea that demonstration of platelet destruction predominantly in the spleen would mean a successful splenectomy and vice versa.^38,39

Alternate concept

The doyen of haematology, William Dameshek suggested in the first edition (1946) of the reputed journal, Blood, that, ‘it would appear that only one conclusion as to the pathogenesis of the disease (ITP) is tenable: i.e., that it is due fundamentally to an abnormality of the spleen, which exerts an unusual effect upon the production of platelets from the megakaryocytes in the marrow’.⁴⁰ In other words, he proposed that a humoral factor is produced by the spleen which can affect the platelet production. In keeping with this fact were the following observations – bone marrow examination within a few hours of splenectomy show platelet budding and within a couple of days, huge masses of platelets being released.⁴⁰ This is noticeable in patients who had splenectomy who almost always show a dramatic increase in platelet count, soon after the surgery, which occurs due to the ‘lack of inhibitory effects’ from the spleen on platelet production.

Further on from the same group of Dameshek, some direct experiments involving splenic circulation was performed.⁴¹ This included identifying the differences in platelet number and function between bloods collected from the splenic artery and splenic vein after injection of platelet-rich blood into the circulation and also directly into the splenic artery. These tests showed no significant differences in platelet number or function confirming the failure of the concept of ‘selective sequestration’ of platelets by the spleen in ITP. Fujisawa et al.⁴² examined the clearance of platelet antibodies in ITP patients who were noted to have an improvement in the platelet count. Splenectomy as much as prednisolone and cyclophosphamide, appeared to act by primarily affecting autoantibody production.

In summary, splenectomy is useful for patients with ITP, not just because, it is site of platelet destruction but also due to the facts that antibody production from the organ and its inhibitory role on platelet production is removed.

Splenectomy increases the risk of infections in patients with ITP

Current concept

It is widely known that splenectomy increases the risk of infections.⁴³ Guidelines have been produced to aid physicians in dealing with such patients and as an aide memoire for vaccinations and prophylactic antibiotics.^44,45 Since the first description of overwhelming post-splenectomy infection published by King and Schumaker in 1952, an estimate incidence of 0.18–0.42% per year, with a lifetime risk of 5% has been quoted for this complication although only limited data relating to morbidity and mortality exist primarily due to lack of systematic studies.^46,47

Alternate concept

Although removal of an important immune organ like spleen is expected to cause an increase in infections from encapsulated organisms, mainly and others, studies from the past has shown that this is restricted to those who have haemolytic diseases and malignancies. For example, a paper published in NEJM in 1967, demonstrated that the risk of infections are higher in those with thalassemia compared to those who had spleen removed for ITP, or tumours.⁴⁸ Two most quoted papers in this regard confirm this finding. Holdsworth et al.⁴⁹ undertook a critical review of the literature on postsplenectomy sepsis from 1952 to 1987. This included a cohort of 12 514 patients although only about 6000 were analyzable. They concluded that severe infection after splenectomy for benign disease is very uncommon except in infants and children below the age of 5 years and that many of the reported postsplenectomy infections may have been coincidental. More recently, Bisharat et al.⁵⁰ reviewed 78 studies published between 1966 and 1996. Of 19 680 patients who had had a splenectomy, the incidence of infection was higher for patients with thalassaemia major and sickle cell anaemia than for patients with ITP.

Specifically looking at patients with ITP and splenectomy infection rates, Schwartz et al.⁵¹ conducted a retrospective study of 75 patients with ITP who underwent splenectomy between 1988 and 1993 at three major medical centres and had a 5-year follow-up. Of the 56 patients (78%) evaluable, no life-threatening infections developed after splenectomy in the 434 patient-years of follow-up. Another retrospective analysis of 402 patients who underwent splenectomy for ITP between 1959 and 2002 in 22 different haematology centres in Europe, infection was not a significant problem.⁵² The most recent publication on this topic analyzed the data of 233 ITP patients, who underwent splenectomy between 1959 and 2001, in six European haematological institutions and had a minimum follow-up of 10 years from surgery.⁵³ Seventy-three (31%) patients experienced at least one infectious complication, which was fatal in two patients. Interestingly, a stable response to splenectomy was associated with a lower rate of infections (P = 0.004). It is tempting to believe that heavy immunosuppression which is tried in refractory cases may be a risk factor for the infectious episodes although it was not possible to be ascertaining this conclusion.

Although the infection risk is often considered, other vascular complications like thrombosis and pulmonary hypertension are often overlooked in patients who have undergone splenectomy.⁵⁴ Both local and systemic venous thromboses have been noted after splenectomy. In the largest report including nearly 9000 patients, 32% of the reported thromboembolic events were deep venous thrombosis, and 13% pulmonary embolism cases, while 16% were portal venous thrombosis.⁵⁵ It has been suggested that the process of pulmonary hypertension may actually represent ‘in situ’ thrombosis.⁵⁶ Due to this high rate of complication, thromboprophylaxis after splenectomy, including anticoagulation or antiplatelet agents, has been recommended.⁵⁴

In summary, although splenectomy is associated with an increased rate of infections, in general, in patients with ITP specifically this risk is minor and is usually noted in refractory cases. More focus should be put on discussing the vascular complications and measures to prevent them.

Conclusion

In conclusion, several established concepts in relation to the role of prednisolone in the management of ITP are engrained in physicians’ minds although alternate explanations may be provided for them. In a similar way, although less often performed at least in the well-resourced settings, splenectomy, still occasionally considered as a treatment of ITP may be effective in removing the organ, which suppresses platelet production. Also, such a surgery is not often associated with an increased risk of infection in ITP patients but with vascular complications.