Immune Thrombocytopenia with Pulmonary Embolism and Deep-Vein Thrombosis: Recommendations for Bone Marrow Aspirate and Biopsy

In an article written for The Hematologist in 2011, I reviewed the results of several small series and population-based analyses that assessed the risk for thromboembolism in patients with immune thrombocytopenia (ITP). I concluded that “these studies provide evidence that the incidence of thromboembolism is increased in patients with ITP, even in the presence of very low platelet counts.” Since that time, there has been no evidence presented to suggest that that conclusion is not accurate. Indeed, several additional analyses involving large numbers of patients with ITP support the conclusion that ITP is indeed associated with a small, but significant risk of thromboembolism,¹  a risk that many hematologists are now aware of.

In most patients with ITP, the increased risk of thrombosis is small (relative risk < 2) and not sufficient to influence management.²  However, more recent studies suggest that certain subpopulations of patients with ITP may be at a significantly higher risk of thrombotic complications. These include patients who have undergone splenectomy; several studies, including a large study that identified patients with ITP undergoing splenectomy through administrative codes,³  suggest that thrombotic risk is increased more than 5-fold in the 90 days after surgery, and 2.7-fold thereafter. Patients older than 60 years are also at increased risk of thromboembolism. Finally, though no definitive data confirm an increased thrombotic risk associated with thrombopoietin receptor agonists, some studies do suggest increased risk with these agents, and this should be considered when treating patients with ITP who have other significant thrombotic risk factors.

(Editor's Note: This question was submitted through the Consult-a-Colleague Program. Dr. McCrae was asked to respond.

How often do you see immune thrombocytopenia (ITP) with pulmonary embolism (PE) and deep-vein thrombosis (DVT)? The patient is a 60-year-old woman who had pneumonia about a month ago and now is presenting with a platelet count of 30,000/μl with DVT and PE. A hypercoagulable diagnostic panel, including tests for heparin-induced thrombocytopenia (HIT) and antiphospholipid antibodies, are pending. Would you perform a bone marrow aspirate and biopsy? A pulmonologist inserted an inferior vena cava filter; was that appropriate?

ITP is a fascinating disorder with a complex pathogenesis that has only recently begun to be understood in detail.^1,2  While a number of immune abnormalities underlie the development of ITP, the disease is largely mediated by antibodies reactive with glycoproteins expressed on platelets and megakaryocytes. These antibodies cause accelerated platelet destruction in the spleen as well as deficient platelet production secondary to impaired maturation of megakaryocytes.

The most common clinical presentation of ITP is bleeding, the severity of which correlates with the degree of thrombocytopenia.³  However, many patients with ITP have few bleeding complications even with very low platelet counts, an observation that may be related to the presence of young, highly functional platelets, increased levels of plasma microparticles, or other factors. Surprisingly, recent studies have also demonstrated that patients with ITP have an increased risk of thrombosis. The data concerning ITP and thrombosis are derived from several sources. A retrospective study from Aledort et al. suggested an increased risk of thrombosis in patients with ITP, with 18 thromboembolic events occurring in 186 adults (the majority after the diagnosis of chronic ITP).⁴  In another retrospective study reported in an abstract by Bennett et al., a 6.9 percent cumulative incidence of thromboembolic events in patients with ITP was seen over a 15-month period. Sarpatwari et al. prospectively analyzed the incidence of thromboembolic events in patients with ITP using the UK General Practice Research Database, identifying 1,070 adults &gt;18 years of age with primary ITP that were matched with 4,280 ITP-free patients.⁵  Over a median of 47.6 months, the adjusted hazard ratio for venous, arterial, or combined thromboembolic events in patients with ITP was 1.58 (95% CI, 1.01-2.48), 1.37 (95% CI, 0.94-2.00), and 1.41 (95% CI, 1.04-1.91), respectively. Subgroup analysis suggested a strikingly increased risk of myocardial infarction in the ITP cohort. Also suggested was a direct relationship between the severity of thrombocytopenia and the risk of thrombosis. Another recent study utilizing a matched ITP cohort obtained from the Danish National Patient Registry observed an incidence rate ratio for venous thromboembolism in patients with ITP of 2.04 (95% CI, 1.45 2.87).⁶  Two of the ITP patients with VTE had severe thrombocytopenia, although one was ultimately diagnosed with lung cancer.

Taken together, these studies provide evidence that the incidence of thromboembolism is increased in patients with ITP, even in the presence of very low platelet counts. While acute portal vein thrombosis has been reported in up to 15 percent of ITP patients undergoing splenectomy, and an increased risk of subsequent thrombosis has been associated with splenectomy for thalassemia,⁷  the review from Sarpatwari et al. demonstrates a similar incidence of thrombosis in splenectomized and nonsplenectomized ITP patients.

The mechanisms underlying the paradoxical development of thrombosis in patients with ITP have not been defined, though several have been postulated. The incidence of antiphospholipid antibodies (APLA) appears to be increased in patients with ITP, and several studies suggest that ITP patients with APLA develop thrombosis more frequently. In a prospective study, Diz-Küçükkaya assessed 82 newly diagnosed patients with ITP, finding 31 (37.8%) positive for APLA.⁸  After five years, the cumulative thrombosis-free survival of APLA-positive and APLA-negative patients was 39 percent and 97.7 percent, respectively (p=0.004). Though current guidelines do not recommend routine screening of ITP patients for APLA,^9,10  this should be considered in those who develop thrombi. Additional factors contributing to the development of thrombosis in patients with ITP may include elevated levels of platelet-derived microparticles¹¹  and complement activation on antibody-coated platelets.¹² 

The management of thrombosis in thrombocytopenic patients with ITP is challenging and not addressed by current guidelines. Likewise, there are no evidence-based data on which to draw from for recommendations. Many experts suggest that the risk of anticoagulation is acceptable at platelet counts above 40,000 to 50,000/μl, although the severity of the thrombotic event and clinical characteristics of the patient should be considered. In the case presented here, the clot burden is extensive and the platelet count close to where anticoagulation is reasonable. Risk factors for bleeding in ITP include a prior history of bleeding and age ≥ 60 years, the latter of which is applicable to this patient. I would recommend aggressive treatment of this patient’s ITP, initially using corticosteroids, since agents such as IVIg may potentially predispose to further thrombosis, and I would consider institution of anticoagulation with intravenous heparin or another short-lived anticoagulant once the platelet count increases to ≥45,000/μl. The use of an IVC filter is a matter of preference; it may be more strongly justified if there is evidence of progressive thrombosis, or if the thrombocytopenia does not respond quickly. However, given the association of ITP with thrombosis, I would avoid intravascular devices if possible, and if an IVC filter is used, I would recommend a removable model.

The decision to perform a bone marrow aspirate and biopsy should not be influenced by a thrombotic event. If the peripheral blood smear shows isolated thrombocytopenia and the history and physical examination do not suggest other disorders, bone marrow aspirate would not be required.