Refractory Primary Immune Thrombocytopenia With Bleeding and Thrombosis: A Case Report

Introduction

Immune thrombocytopenia (ITP) is an autoimmune disease characterized by a peripheral blood platelet count below 100×109/L and based on increased antibody-mediated platelet destruction along with reduced thrombocytopoiesis, due to an impaired immune-mediated megakaryocytic process [1–3]. In particular, underlying the pathophysiology of ITP is a dysfunction of the immune system that affects multiple cellular components. In ITP, there is a dysfunction of T helper (Th) cells, which directly stimulate B cells generating autoantibodies as well as a dysregulation of regulatory T cells, and this dysregulation suppresses the generation of regulatory cells controlling autoantibodies. The autoantibodies in question are immunoglobulin G (IgG) antibodies directed against GPIIb/IIIa and GPIb of the platelet membrane, causing platelet clearance from the circulation through the reticulo-endothelial system of the liver and spleen, leading to thrombocytopenia. The anti-GPIb antibodies cause the loss of sialic acid in platelets, accelerating the process of platelet clearance. Platelet autoantibodies also bind to megakaryocytes expressing GPIIb/IIIa and GPIb, causing production and maturation suppression, resulting in thrombocytopenia. In addition, it has been observed that cytotoxic T cells cause direct platelet and megakaryocyte lysis [4].The disease of ITP is characterized mainly by bleeding manifestations of the skin and mucous membranes [4], especially when platelet count is less than 20×109/L, although possibly fatal hemorrhage of the retroperitoneal and intracranial space are the most severe manifestations[5,6]. It is also worth mentioning that risk of bleeding in ITP may partially be attributed to platelet dysfunction [7].On the other hand, antibody complexes have been described to activate both platelet function and the complement system, among other mechanisms, and together with activated megakaryocytes can theoretically induce overexpression of coagulation factors, increasing the risk of thrombosis [1]. Moreover, risk factors, such as obesity, hypertension, diabetes, or a history of positive antiphospholipid antibodies or cardiovascular diseases, further enhance the risk of thrombotic events [8]. Current ITP treatment algorithms following first-line therapy are not strictly regimented, as later lines consist of different drugs such as thrombopoietin receptor agonists (TPO-RAs), namely eltrombopag, romiplostim, and avatrombopag; the monoclonal antibody rituximab; the spleen tyrosine kinase (Syk) inhibitor fostamatinib; immunosuppressant therapy (mycophenolate mofetil, cyclosporine, azathioprine, danazol, and dapsone); and splenectomy [9–12]. Some of these therapies, in particular RAs, were also identified as potential risk factors for thrombotic events [9]. Currently, there are no specific, detailed therapeutic guidelines for managing and balancing bleeding and thrombotic state in an ITP patient, and in case of thrombotic events, the most important thing is achieving sufficient platelet count to be able to treat the patient with full-dose anticoagulant or antiplatelet drugs [9]. Here, we describe an unusual case of refractory ITP associated with both major bleeding and thrombotic events following first-line treatments and subsequent treatments, and discuss how challenging it can be to individualize and tailor therapy in these extremely complicated cases. This case provides more comprehensive details for improved patient selection for specific ITP therapy and necessary prevention of thrombosis. In fact, in this case report, the patient was finally treated with avatrombopag in association with dual antiplatelet therapy, and showed a good response in terms of platelet count and thrombosis prevention.

Case Report

A 50-year-old male patient was diagnosed with ITP in year 2000, with platelet counts below 30×109/L. The diagnosis was confirmed by exclusion with bone marrow study negative for hematological tumors, absence of secondary causes of ITP, other than gastritis positive for Helicobacter pylori, and presence of anti-platelet antibodies. In past medical history, a previous episode of cerebral ischemia 1year before ITP onset and his smoking habit were highlighted. Given the lack of response in terms of platelet count following H. pylori eradication and worsening of thrombocytopenia, the patient was started on corticosteroid therapy with oral prednisone 1 mg/kg daily for 3 weeks, achieving only a weak partial response (count between 20–30×109/L), followed by prolonged low-dose tapering. After a petechial gastritis bleeding episode in October 2002, the patient refused further oral corticosteroid therapy and was treated periodically with intravenous immunoglobulins (IVIGs) 1 g/kg daily for 2 consecutive days, and occasional platelet transfusions together with tranexamic acid in case of persistent bleeding episodes. It should be mentioned that between 2003 and 2007 the patient was also treated with vincristine 2 mg IV chemotherapy once a week for 4 weeks and anti-D immunoglobulin (50 μg/kg in a single dose) without any benefit, while dexamethasone 40 mg IV for 4 consecutive days was used only in extreme cases with persistent bleeding. In April 2008 the patient suffered from cerebral hemorrhage and, given an extremely high bleeding profile, splenectomy was performed in June 2008 followed by rituximab immunotherapy (375 mg/m2 weekly for 4 weeks), achieving only partial response, with a platelet count of 35×109/L. The platelet response was lost 1 year after treatment completion and the patient continued with periodic cycles of high-dose IVIGs in case of persistent bleeding events. Following the introduction of TPO-RAs in everyday practice, the patient was started on romiplostim treatment in July 2011 (initial dose 1 μg/kg weekly, subcutaneously). However, this was promptly discontinued following the first dose, due to the occurrence of myocardial infarction treated with coronary stents and subsequent treatment with low-dose acetylsalicylic acid (ASA) and clopidogrel. In the following years, the patient was treated when clinically needed with plateletpheresis, corticosteroids, and IVIGs, while immunosuppression with azathioprine and mycophenolate mofetil did not induce any response. Given the lack of response and no alternatives, in June 2016 he was started on eltrombopag, independently from the regimen for a previous thrombotic cardiac episode, at a lower dose (25 mg daily oral), that was later increased; unfortunately without benefit. Four months later, rituximab therapy was tried once again, however without response. One year later, following hospital approval, bortezomib was used off-label at dose 1.3 mg/m2 twice weekly for 2 consecutive weeks, receiving a total of 4 cycles, without response. Dapsone was also tried, at a dose of 75 mg in January 2018 with no effect in terms of platelet count. Given the lack of alternatives in March 2019, romiplostim was re-started, resulting in a significantly increased platelet count (358×109/L) at a dose of 2 μg/kg. Unfortunately, 2 months later, the therapeutic response was lost and the romiplostim dose was increased to 6 μg/kg, after which the patient suffered from acute coronary syndrome and was prescribed again low-dose ASA without romiplostim suspension. In October 2019, the patient suffered from unstable angina pectoris, and a percutaneous transluminal coronary angioplasty (PTCA) was performed with stent placement, followed by dual anti-platelet therapy. Due to significant platelet count fluctuations and in the absence of alternatives, the patient continued personalized romiplostim therapy (4 μg/kg every 10–15 days), until June 2022, when fostamatinib became available. He was started on 100 mg twice daily oral therapy, but the therapy was quickly stopped due to the onset of melena, and romiplostim was re-administered once again. In January 2023, the patient switched to avatrombopag (20 mg daily) but continued to suffer from important platelet fluctuations up to 1.0×109/L. Treatment was then personalized according to the value of platelet count. In January 2024, the patient developed an encephalopathy with hypoacusia and impairment of gait, probably due to an important immune deficit caused by different immunosuppressive treatments, including rituximab, and previous splenectomy. He remains on avatrombopag 20 mg daily therapy in association with dual antiplatelet therapy with good platelet response (219×109/L in December 2024). In Figure 1, a diagram on this case is reported.

Discussion

ITP is a primary autoimmune disorder characterized by a platelet count <100×109/L[13], with annual incidence ranging around 2–7 cases per 100 000 inhabitants [14]. It is more common in adults than children [15,16]. Mortality related to the disease and its complications seems to be 1.3–2.2 times higher compared with the normal population [17].

Thrombotic events represent a paradoxical and unexpected complication in ITP given the low platelet count [18]. However, some retrospective cohort studies [19,20] have shown that the risk of arterial and venous thromboembolic events is higher in ITP patients than in the normal population. The pathophysiology of the thrombosis in ITP is not clearly understood. Immature hyperactive platelets, procoagulant microparticles, procoagulant megakaryocytes expressing von Willebrand Factor (VWF) and coagulation factors, and endothelial damage releasing VWF, could all be responsible for thrombus formation in ITP [1]. Moreover, some ITP treatments and previous splenectomy potentially increase the risk of thrombosis [21].

Other acquired risk factors for thrombosis independently from ITP are advanced age, obesity, hypertension, diabetes, previous thrombotic history, and antiphospholipid antibodies [6]. Smoking habit is another important risk factor, since it promotes the adhesion of monocytes/lymphocytes and platelets to the inflammatory endothelium. This could lead to an increase of VWF levels together with platelet adhesion and thrombus formation [22–24].

Previous splenectomy is a well-known risk factor for thrombosis, with a 5-year cumulative incidence of venous and arterial thrombotic events of 6.6% and 10.2% respectively [25]. The splenectomy-related thrombosis data differ in different studies and this variation has meant that there are is no fixed consensus on the subject [26]. A possible explanation could be the lack of clearance of thrombogenic microparticles following splenectomy. Registry studies in California showed that the risk of vein thrombosis was 4.3% vs 1.7% with and without splenectomy, respectively [27].

Certain drugs used for the treatment of ITP have also been indicated as possible risk factors for the development of thrombosis, mainly TPO-RAs. The incidence of thrombotic events, both arterial and venous, was around 6% in ITP patients treated with TPO-RAs after a median follow-up of 2 years, especially when associated with other known risk factors [28,29]. The pathogenesis of thromboembolic events during TPO-RA therapy, however, is not clear. Furthermore, it seems that there is no correlation with high platelet counts caused by the treatment, given that thrombotic episodes were described even in cases of partial or no response [10].

The importance of taking into account the above-mentioned risk factors for thromboembolism, when an ITP patient has to be medically treated, is well described in our case reported here. Our patient had a previous episode of cerebral ischemia, 1 year prior to ITP onset, together with smoking habit; all conditions that even by themselves predispose patients to a higher risk of thrombosis. Moreover, the patient underwent splenectomy first, and suffered from myocardial infarction following the first administered dose of romiplostim therapy. Furthermore, after romiplostim was re-started due to lack of alternatives, the patient had ulterior cardiac complications, including acute coronary syndrome, and unstable angina pectoris followed by PTCA.

Treating ITP patients can be somewhat challenging in certain cases. American Society of Hematology (ASH) guidelines from 2019 recommend starting therapy when platelets are below 20–30×109/L. The first-line treatment is based on corticosteroids or IVIGs, if corticosteroids are contraindicated [30], which should also include anti-D immunoglobulin [31]. In case of loss of response to first-line treatment, the recommended second-line therapies are TPO-RAs and rituximab, while fostamatinib and splenectomy are suggested for chronic ITP patients [10].

The introduction of TPO-RAs has been revolutionary in terms of platelet response in ITP, binding to the TPO receptor with the activation of the JAK-STAT pathway, resulting in an increase of megakaryocyte progenitors and platelet production in the bone marrow [10]. The efficacy of romiplostim and eltrombopag has been well documented in several studies [28–30,32]. Two parallel phase III studies carried out in Europe and the USA [32] showed platelet response rates in 88% of non-splenectomized and 79% of splenectomized patients treated with romiplostim vs 14% of non-splenectomized and 0% of splenectomized patients who received placebo. In addition, a good response has been achieved with the new TPO-RA avatrombopag in the treatment of chronic ITP patients. In some retrospective studies [33], there was a surprisingly high rate of response with avatrombopag, when switched from both eltrombopag and romiplostim. However, as mentioned above, thromboembolism is a worrying adverse effect for all 3 drugs [34].

Although the recommendations of the 2019 ASH guidelines do not address the issue in a specific manner, treatment of ITP patients with high risk for thrombosis or history of previous thromboembolic episodes can be challenging and should be individualized. Solid scientific evidence is not available and the choice is based mainly on expert opinions and individual patient cases. Our case could be considered as an example of how difficult it can be to balance bleeding and thrombotic risks, especially when the patient is unresponsive to alternative treatments. The initial clinical course in this case was characterized by bleeding events and, therefore, the therapy consisted of corticosteroids administration, which was stopped when a bleeding gastritis occurred, and was followed by “on demand” platelet transfusion and periodical cycles of IVIGs. After the onset of a major bleeding event, cerebral hemorrhage due to persistently low platelet count, splenectomy was performed, followed by rituximab, achieving only a partial transitory response.

The cardiovascular thrombotic events that occurred in our patient while taking romiplostim (myocardial infarction at first, and both acute coronary syndrome and unstable angina pectoris later), which were treated twice with PTCA, should have been sufficient to discontinue the drug permanently. However, due to the necessity of dual antiplatelet therapy, and complete absence of response to treatment alternatives such as azathioprine, mycophenolate mofetil, bortezomib, rituximab, dapsone, and low-dose eltrombopag, we had no alternatives other than to continue romiplostim treatment, given the importance of maintaining normal platelet counts to maintain adequate antithrombotic treatment. Another aspect that should not be underestimated is frequent fluctuations of the platelet counts attributed to TPO-RAs, leading to the requirement for a personalized treatment regimen in terms of drug administration dose/frequency. An attempt was made to switch to fostamatinib, an oral Syk inhibitor with a significantly lower thrombotic risk [10], as soon as it became available. However, the patient suffered from melena, and was forced to continue TPO-RA treatment, first with romiplostim once again, and later with avatrombopag. With a tailored dose of avatrombopag (in association with dual antiplatelet therapy), our patient showed a good persistent response with platelet counts higher than 200×109/L, although periodical fluctuations remained. Having achieved both good clinical and laboratory response, the next challenge was to progressively discontinue the therapy with avatrombopag, given the known thrombotic risk [34]. Data in the literature [35,36] describe the possibility of treatment discontinuation, prior to dose tapering, and maintenance of “sustained remission off-treatment” in 10–30% of patients. However, it was described only for eltrombopag and romiplostim, given their immunomodulatory effect, and it remains unclear how long the treatment should be continued after the patient has achieved stable platelet levels, before dose tapering should be attempted. In our case, a history of major bleeding and the necessity of high platelet levels raise many concerns, and therefore, the safest course of action is probably to continue the therapy unchanged, until novel agents become available.

There are different drugs under investigation with various mechanisms of action, from blocking the production of circulating autoantibodies, stimulating their removal, and interfering with the complement immune response, to splenic and hepatic platelet clearance reduction [37–40]. These drugs will hopefully be able to improve the response rates in refractory chronic ITP along with reduced thrombotic risk. Among these new drugs, rilzabrutinib, a potent BTK inhibitor, seems to have a long-lasting effect, according to the report of Kuter et al [41], and, interestingly, no patient thrombotic events of grade 2 or higher [41] were described in that report.

Conclusions

Our clinical case highlights the difficulty of managing patients with multi-resistant ITP and dual bleeding and thrombotic risk. Lacking international guidelines, our report suggests a collaborative approach between patient and clinician about care and risk-based and individualized management for bleeding and thrombosis prevention/treatment in adults with primary ITP.